TITLE
A Novel Kiosk System for Flat Foot Analysis and Foot Morphology Using AIBased Image Processing and its Method thereof
[0002] FIELD OF THE INVENTION
[0003] The present invention relates to biomechanical foot analysis devices and,
more particularly, to a self-operated novel kiosk system and method for the
detection of flat feet (pes planus) and comprehensive foot analysis. The
system utilizes deep learning based image segmentation and computer vision
techniques to not only detect flatfoot conditions but also to determine arch
height, instep height, arch type, and foot dimensions with high accuracy,
following its novel method.
[0004] BACKGROUND OF THE INVENTION
[0005] Flat feet, or pes planus, is a condition characterized by the collapse of the
medial longitudinal arch of the foot, often resulting in full or near-full contact
of the sole with the ground. This condition can lead to pain, altered
biomechanics, gait abnormalities, and other musculoskeletal complications.
Existing diagnostic methods typically rely on expensive equipment operated
by trained professionals, making them inaccessible for routine or mass
screening. Furthermore, current foot measurement systems used for
applications such as shoe fitting are often separate from clinical diagnostic
tools.
[0006] There exist several foot measurement systems, each with varying levels of
accuracy, cost, and applicability. Some of the existing systems are discussed
below.
[0007] Brannock Device (Manual Sizer)- A mechanical tool commonly used in retail
footwear stores to manually measure foot length and width. While simple and
inexpensive, it lacks the ability to assess arch type, instep height, or overall
foot structure.
[0008] Footprint-Based Methods (Ink or Digital Scans) - Techniques like the Staheli
Index or Clarke’s Angle are derived from 2D footprint images or ink
impressions. These are low-cost but often inaccurate and subjective,
especially when diagnosing flatfoot conditions.
Page 2 of 30
[0009] 3D Foot Scanners - High-end systems using structured light or laser scanning
to generate 3D foot models. These are highly accurate but expensive, bulky,
require trained personnel, and are mostly used in orthopaedic clinics or
premium footwear brands, making them impractical for mass screening or
use in rural/low-resource settings
[0010] Pressure Mats and Podoscopes – These devices analyze plantar pressure
distribution and arch loading. While useful for dynamic gait analysis, they
cannot capture side profile parameters like arch height or instep height and
are often costly.
[0011] Smartphone Photogrammetry Tools - Emerging mobile applications use phone
cameras and image stitching to estimate foot size and shape. However, they
often lack clinical accuracy, require user skill, and don’t offer comprehensive
foot morphology data.
[0012] Further the existing systems are quite expensive. 3D foot scanners
(structured light or laser-based) costs around ₹5–15 lakhs per unit, Pressure
mat systems costs around ₹2–8 lakhs depending on resolution and brand,
Gait labs with motion capture and force plates costs ₹20–50 lakhs, while
Podoscopes and baropodometric platforms costs around ₹1.5–4 lakhs.
[0013] These systems also require trained operators, dedicated clinic space, and
ongoing maintenance, increasing the total cost of ownership. Conventional
methods rely on manual observation, footprint analysis, or ruler-based
measurement, which are subjective, error-prone, and non-repeatable. Most
existing solutions either focus on diagnostics (e.g., arch classification) or on
sizing (e.g., for footwear), but not both. This results in fragmented workflows
and added cost. The inability to generate actionable, accurate foot metrics
hinders the design of custom insoles or corrective footwear, especially in retail
environments.
[0014] Studies confirming growing interest in low-cost, accessible, and automated
flatfoot detection, especially in India is evident from various literature and
study, such as
a. Kalghatgi et al. (2025)- “Flatfoot Detection in an Indian Population: Validation
of Morphological Indices Using a Diagnostic Device” - Eng. Proc. 95(1):6,
June 2025 - This study presents a low-cost system using a transparent glass
Page 3 of 30
platform, LED backlighting, and webcam images to compute traditional
flatfoot indices (Staheli Index, Clarke’s Angle, etc.) using a Python application.
ArUco markers were used for calibration and repeatability. While useful, the
system relies on footprint-based indirect indices and lacks true contour
segmentation or side-view analysis.
b. Gambhir et al. (2023) - “Flat Feet Diagnosis Tool via Photogrammetry and
ML” - This work explores 3D foot modeling using mobile-based
photogrammetry and ML. Although innovative in using smartphones, it lacks
real-time analysis, arch classification accuracy data, and practical deployment
results.
[0015] However, they remain limited in dimensional accuracy, side profile analysis,
and complete automation.
[0016] There are several patent which works in this field but have one or more
limitations which are to be addressed. Some prior arts are as follows:
[0017] CN113057623B – Flat Foot Screening Gait Analysis System -Describes a
pressure-sensor insole using K-NN machine learning to classify gait data.
Focuses on dynamic gait and plantar pressure rather than static image-based
structural analysis.
[0018] CN102184539A & CN102184539B – Image-Processing Based Plane Footprint
Detection System -Describes depth analysis from footprints and use of image
processing for forensic and biomechanical insight. Although technically
robust, these methods are 2D footprint-dependent and do not use deep
learning segmentation or side-view profile evaluation.
[0019] US11816806B2 – System for Foot Scanning via Mobile Computing Device -
Involves generation of 3D models of the foot using mobile cameras to create
point clouds. Focus is on morphable 3D model generation, not automated
classification or kiosk-based use.
[0020] US20210093050A1 – Method for Measuring Foot Size and Shape Using Image
Processing - Describes capturing a foot image alongside a reference object
and extracting foot dimensions. This system is static and manual in nature,
lacking deep learning or full morphological classification.
Page 4 of 30
[0021] There is a need for a scalable, automated, and cost-effective solution that not
only detects flat feet but also accurately measures arch height, instep height,
arch type, and overall foot dimensions. These detailed biomechanical insights
can be leveraged to design and manufacture custom orthopaedic insoles and
corrective footwear, making the system valuable for both clinical diagnostics
and personalized orthotic applications.
[0022] SUMMARY OF THE INVENTION:
[0023] The present invention discloses a self-operating diagnostic kiosk designed to
detect flat feet (pes planus) and perform comprehensive foot analysis using
a U-Net-based image segmentation algorithm and advanced computer vision
techniques. The system is capable of accurately determining foot arch type,
arch height, instep height, and foot dimensions, thereby enabling both clinical
assessment and custom insole design.
[0024] The present invention discloses a novel Kiosk System [800] for Flat Foot
Analysis and Foot Morphology using AI Based Image Processing. The system
comprising a foot placement platform [400] provided with stopper and
boundaries, and calibrated with Aruco marker for spatial accuracy; multiple
high resolution camera modules for capturing images; a processing unit
employing a U-Net-based deep learning model algorithm to compute
dimensional parameters of foot and classify foot arch types, wherein
dimensional parameters are medial arch height, instep height, foot
dimensions; a display interface [803] controlled by the processing unit and
integrated with infrared touch screen panel for interactive user input and
result visualization; and lighting units controlled by a control module
comprising MOSFET based LED controller to regulate LED operation based on
system based commands, all positioned and operatively connected to a main
body disposed with a central mounting frame, curved side panels [603] on
either side of central mounting frame, background shield panel or side optical
shield [601], adjustable camera mount bracket [200], a display mount stand
[100], cables and connecting accessories.
[0025] DETAILED DESCRIPTION OF FIGURES
[0026] Figure 1- The figure 1 illustrates the extended arm component of the device,
which serves as the structural support for the display unit. The arm's length
determines the height at which the screen is positioned, and this has been
carefully designed to ensure ergonomic visibility and ease of use for most
Page 5 of 30
users interacting with the kiosk. The top surface of the arm provides a stable
mounting platform for securely holding the display. With a height of 1100
mm, the screen is elevated to an optimal viewing level, allowing users to
comfortably read and interact with the interface while standing, without the
need to bend or strain.
[0027] Figure 2 - Adjustable Camera Mount Brackets - Modular holders that securely
mount camera modules in top-view and side-view configurations. They are
designed to be adjustable, allowing precise alignment of image sensors
relative to the foot placement area for optimal field of view and focus.
Illustrate the structural design and dimensional features of the side-mounted
arm component that holds the cameras and LED lighting used for foot
imaging.
[0028] Figure 2(A) shows the side view of the curved arm. This arm supports the
overhead camera and LED lighting module positioned to capture top-view
images. The diagram indicates that the horizontal distance from the front
edge of the arm to the camera mounting slot is 19.33 mm. This ensures that
the camera is properly recessed from the edge for both protection and optimal
field of view.
[0029] Figure 2(B) presents the front view of the same side arm. It shows the camera
mounting slots located at the centre of the vertical face. The distance from
the bottom of the arm base to the camera position is 140.21 mm, ensuring
side-view cameras are positioned at the correct anatomical height relative to
the foot. The dimension labelled as 37.60 mm indicates the horizontal
distance between the two camera mounting holes on the front face of the
arm. The dimension marked as 24.00 mm represents the thickness of the
vertical strip or body of the arm used in the assembly. This thickness ensures
adequate structural rigidity to support the mounted components such as
cameras and LED modules.
[0030] Figure 2(C) displays the top view of the upper section of the arm, where the
LED lighting module is housed. The drawing shows the distance from the edge
of the arm to the first LED light as 46.02 mm, and the distance between the
two LED lights is 52.50 mm. These measurements ensure symmetrical light
distribution across the imaging field, minimizing shadows and enhancing
image clarity during segmentation and measurement.
Page 6 of 30
[0031] Figure 3 illustrates the camera mounting bracket, which securely holds the
imaging module and attaches directly to the side arm structure shown in
Figure 02. The bracket is engineered for precise positioning and alignment of
the camera, ensuring optimal image capture from both lateral and top views
of the foot.
[0032] Figure. 3A: This is the front view of the camera mount. The dimension marked
as 15.70 mm represents the distance from the top edge of the mount to the
starting point of the camera mounting holes. The smaller circular holes denote
the camera mounting slots, with the screw holes separated vertically by 4.13
mm and spaced 4.00 mm apart horizontally, allowing compatibility with
standard camera modules. The larger outer holes represent the mounting
points used to affix the bracket securely to the side-arm structure described
in Figure 2(B).
[0033] Figure. 3B: This is the side view of the mount, showing the camera tilt angle.
The camera is fixed at an inclination of 3.50°, which has been intentionally
designed to align the lens with the anatomical arch of the human foot. This
slight angular positioning improves visibility and accuracy in arch detection
and side-profile analysis, particularly in identifying conditions such as flatfoot
or high arch.
[0034] Figure 3C: This is the isometric or complete 3D view of the camera mount,
providing a holistic visual representation of its geometry, mounting interface,
and structural features. It shows how the mount interfaces with both the
camera module and the arm assembly, ensuring secure fitment and consistent
orientation.
[0035] Figure 4 - Foot Placement Platform - This is the flat base where users position
their feet for top and side image acquisition. It includes calibrated reference
markers or ArUco tags to enable accurate millimeter-per-pixel conversion.
The platform ensures consistent alignment and placement for reliable
diagnostic results. Illustrates the base plate of the kiosk, which serves as the
designated foot placement area for users during scanning. The layout and
dimensions of this base have been carefully designed for ergonomic
alignment, ease of use, and precise image capture. The total width of the base
plate is 550 mm, which accommodates a natural and comfortable standing
Page 7 of 30
position for most users. The distance between the designated positions for
the left and right foot is 310 mm, representing a standard shoulder-width
stance that ensures balance and consistency during the diagnostic process. A
heel stopper is mounted at a distance of 60 mm from the rear edge of the
base, aligned with the position of the side-facing cameras. This ensures that
the user’s heel is consistently placed for repeatable foot positioning and
optimal lateral image capture. Furthermore, there is a 120 mm gap between
the outer edge of the foot placement area and the inner surface of the sidearm structure. This spacing is calculated to provide the necessary clearance
for the side cameras to capture unobstructed lateral images of both feet, while
also ensuring that lighting and segmentation accuracy remain consistent
across users. The base plate’s markings and structural elements work in
unison with the kiosk's imaging components to provide a standardized,
comfortable, and precise setup for every diagnostic session.
[0036] Figure 5 - It provides multiple views of the base structure of the kiosk,
highlighting the key mounting dimensions and support features designed for
stability and modular attachment of other components.
[0037] Figure 5A: This is the top view of the kiosk base. It shows two metal support
strips mounted on the base at a distance of 242.60 mm from each other.
These supports are strategically positioned to provide a firm resting platform
for the user's feet, ensuring correct alignment during the scanning process.
[0038] Figure 5B: This is the side view of the base. It illustrates the overall height of
the base platform, which is 50 mm, including the raised section where the
display arm (as shown in Fig. 01) is mounted. The height of the mounting
area for the side-arms is marked as 22 mm, and the width of the same
mounting area is 40 mm, matching the dimensions required for securely fixing
the side arm assemblies (as seen in Figure 02).
[0039] Figure 5C: This is the front view of the base. It highlights the mounting
holes for attaching the arm structures.
[0040] Figure 6 - Arm Cover Plate / Decorative Side Cover - These are external
enclosures mounted over the side curved panels, primarily for cosmetic
finishing and to protect internal cabling and components such as LEDs and
camera mounts. They also prevent environmental dust from entering the
sensor compartments. Illustrates the Background Shield Panel, also referred
Page 8 of 30
to as the Side Optical Shield, which is vertically mounted alongside the foot
placement platform of the kiosk. The primary function of this panel is to block
ambient background elements during image acquisition, thereby enhancing
image contrast and segmentation accuracy, especially in environments with
uncontrolled lighting or visual clutter. Background Shield Panel / Side Optical
Shield - These panels are mounted vertically alongside the foot placement
platform to obstruct the background during image capture. Their primary
purpose is to ensure high-contrast and background-free images for accurate
contour extraction, especially in uncontrolled lighting environments. The
panel has been precisely dimensioned to align with the height and viewing
angles of the side cameras:
[0041] The total height of the panel is 220.00 mm, providing full vertical coverage to
the lateral imaging field.
[0042] The width of the panel is 558.00 mm, ensuring sufficient horizontal shielding
between the foot and the background environment.
[0043] The depth/curvature extension from the mounting edge is 84.00 mm, forming
a smooth curved profile that reduces internal reflections and maintains optical
clarity.
Additionally, the top corners are rounded with a radius of 40.00 mm (R40.00),
giving the structure a soft, user-safe edge while also maintaining a
streamlined appearance.
[0044] Figure 7 - I/O Port Access Cover / Connector Access Panel - A rear or sidefacing panel that provides user and technician access to essential external
ports such as USB, LAN, or power inputs. It offers protection while allowing
easy connectivity for maintenance, data transfer, or firmware updates.
Illustrates the I/O Port Access Cover, a precision-machined panel designed to
house and protect essential connectivity interfaces of the kiosk system. This
panel is typically mounted on the rear or side surface of the kiosk enclosure
and facilitates direct access to various I/O ports for both users and service
technicians. The overall length of the panel is 228.00 mm, and the height is
46.00 mm, with a uniform thickness of 4.60 mm. The panel features multiple
cutouts and mounting points, carefully aligned and dimensioned for standard
connectors:
Page 9 of 30
[0045] A series of square and rectangular cutouts are positioned along the center
axis of the panel, allowing access to embedded USB, LAN (RJ45), HDMI, and
power supply ports.
[0046] The central rectangular cutout (178.00 mm from the left edge) is typically
used for main-board or power jack access.
[0047] The panel height at the active interface zone is maintained at 38.00 mm,
ensuring alignment with internal port connectors.
[0048] Circular holes located on either side of the cutouts allow for secure mounting
of the panel to the kiosk frame using screws.
[0049] Figure 8 - Display Mount Stand / Monitor Support Structure - This vertical
stand supports and secures the display screen or touch interface. It provides
ergonomic screen positioning for user interaction, result visualization, and
system navigation during the diagnostic process. Central Enclosure / Main
Housing - The main structural body that houses the core electronics,
processing unit, internal cabling, and wiring. It acts as the central mounting
frame for all mechanical and electronic subcomponents, maintaining
structural integrity and protecting internal modules.
[0050] Figure 9- System Assembly. It presents a top-down perspective view of the
fully assembled kiosk system. This includes the core components such as
the CPU housing, foot placement area, side arms, camera modules, and the
touch-enabled display unit. Clearly labelled elements include:
[0051] LAN Cable Port, HDMI Port, and Power Supply Port on the rear side of the
kiosk base—these ports are aligned with the connector access panel (as
described in Figure 6).
[0052] Both side-facing cameras are shown embedded in the side arms, capturing
medial and lateral foot profiles.
[0053] The display support arm is shown securely mounted and leading up to the
main interactive display.
[0054] This figure is essential to understand how external connectivity, power input,
and data interfacing are managed within the compact form factor of the kiosk
Page 10 of 30
[0055] The assembly clearly shows:
[0056] Side cameras mounted on the curved arms, positioned to capture the lateral
foot profile.
[0057] Top cameras and LED holders, visible in the zoomed Detail A, positioned
above the foot platform to capture high-resolution overhead images.
[0058] The main display unit is mounted on a curved support arm, ergonomically
angled for user interaction.
[0059] Connecting wires are routed through internal compartments of the base to
maintain a clean external design and minimize user exposure to cabling.
[0060] This view highlights how each component integrates into the structural layout
of the kiosk to support simultaneous foot scanning and interactive diagnostics.
[0061] Figure 10 – Blown-up image of Figure 9
[0062] Figure 11 – System components
[0063] Figure 12 - Flowchart of the process illustrates the rear perspective of the
assembled kiosk, focusing on the connectivity and interfacing components.
[0064] DETAILED DESCRIPTION
[0065] The present invention builds upon the growing body of research in the field of
automated flatfoot detection and image-based foot diagnostics. Several prior
studies and patent documents have proposed related approaches; however,
none provide a unified, self-operated, deep-learning-based dual-view system
with the features and scalability offered by the current invention.
[0066] The present invention discloses a self-operating diagnostic kiosk designed to
detect flat feet (pes planus) and perform comprehensive foot analysis using
a U-Net-based image segmentation algorithm and advanced computer vision
techniques. The system is capable of accurately determining foot arch type,
arch height, instep height, and foot dimensions, thereby enabling both clinical
assessment and custom insole design.
[0067] The motivation for developing this invention stemmed from a combination of
clinical necessity, market gaps, and the desire to create an accessible, costeffective, and intelligent solution for foot diagnostics. Flat feet (pes planus) is
a highly prevalent condition that often goes undiagnosed, especially in
Page 11 of 30
developing regions, due to the high cost, manual nature, and inaccessibility
of existing diagnostic tools.
[0068] The following table provides a detailed list of all mechanical components used
in the assembly of the self-operating diagnostic kiosk. Each part is uniquely
identified by a part number and includes the corresponding quantity required
for one complete unit.

[0069] The system comprises the following components -
[0070] Mini Computer equipped with an Intel Core i5 processor and 8 GB of RAM,
responsible for executing the image processing algorithms and managing
the user interface.
[0071] Display Unit, consisting of a 24-inch LED monitor integrated with an infrared
(IR) touchscreen panel, enabling interactive user input and result
visualization.
[0072] Camera Modules, specifically four (4) units of 2 Megapixel Full HD USB
cameras, configured for capturing multiple angles of the subject’s feet for
accurate image segmentation and measurement.
[0073] LED Indicators, comprising two (2) high-brightness LEDs, used for
illumination or status indication during the scanning and analysis process.
[0074] Cabling and Connectivity Accessories, including:
– One (1) HDMI cable for video output
– One (1) USB cable for peripheral connectivity
– One (1) USB extender for extended camera reach
– One (1) LAN extender to support network connection in constrained
spaces
[0075] Custom Structural Components, including:
– One (1) 3D printed heel stopper to ensure proper foot placement during
image acquisition
– One (1) ArUco marker printed on vinyl, used for camera calibration and
spatial reference
[0076] Control Module, comprising one (1) MOSFET-based LED controller to
regulate LED operation based on system commands.
Page 12 of 30
[0077] Miscellaneous Components, such as screws, bolts, electrical wires, cable
ties, mounting clips, and adhesives required for assembling and securing the
various hardware components within the kiosk structure.
[0078] In a preferred embodiment the present system involves deploying the
diagnostic kiosk in a standalone, self-service configuration within clinical,
retail, or public environments. The user approaches the kiosk and follows onscreen instructions via the integrated touch display. The system components
are as follows:
[0079] Foot Placement: The user places one foot at a time on the foot placement
platform, aligning the heel with the provided stoppers and boundaries. This is
the flat base where users position their feet for top and side image acquisition.
It includes calibrated reference markers or ArUco tags to enable accurate
millimeter-per-pixel conversion. The platform ensures consistent alignment
and placement for reliable diagnostic results.
[0080] Central Enclosure / Main Housing - The main structural body that houses the
core electronics, processing unit, internal cabling, and wiring. It acts as the
central mounting frame for all mechanical and electronic subcomponents,
maintaining structural integrity and protecting internal modules.
[0081] Side Curved Panel with Camera & LED Mounts - These are curved structural
panels attached to either side of the main body, designed to house side-facing
Name Part No. Quantity
Foot Placement Platform 01 1 pcs
Main Kiosk Frame 02 1 pcs
Side Curved Panel with Camera &
LED Mounts
03 2 pcs
Side Panel Cover 04 2 pcs
Background Shield Panel 05 2 pcs
Connector Access Panel 06 1 pcs
Adjustable Camera Mount Brackets 07 4 pcs
Display Mount Stand 08 1 pcs
Assembly View 09 –
Page 13 of 30
cameras and illumination LEDs. Their ergonomic shape allows optimal
positioning of sensors and contributes to aesthetic and functional symmetry.
[0082] Arm Cover Plate / Decorative Side Cover - These are external enclosures
mounted over the side curved panels, primarily for cosmetic finishing and to
protect internal cabling and components such as LEDs and camera mounts.
They also prevent environmental dust from entering the sensor
compartments.
[0083] Background Shield Panel / Side Optical Shield - These panels are mounted
vertically alongside the foot placement platform to obstruct the background
during image capture. Their primary purpose is to ensure high-contrast and
background-free images for accurate contour extraction, especially in
uncontrolled lighting environments.
[0084] I/O Port Access Cover / Connector Access Panel - A rear or side-facing panel
that provides user and technician access to essential external ports such as
USB, LAN, or power inputs. It offers protection while allowing easy
connectivity for maintenance, data transfer, or firmware updates.
[0085] Adjustable Camera Mount Brackets - Modular holders that securely mount
camera modules in top-view and side-view configurations. They are designed
to be adjustable, allowing precise alignment of image sensors relative to the
foot placement area for optimal field of view and focus.
[0086] Display Mount Stand / Monitor Support Structure - This vertical stand
supports and secures the display screen or touch interface. It provides
ergonomic screen positioning for user interaction, result visualization, and
system navigation during the diagnostic process.
[0087] The process of the system is as follows:
[0088] Automated Image Capture- The system captures top-view and side-view
images using synchronized high-resolution cameras. Built-in LED lighting
ensures consistent illumination.
[0089] Real-Time Processing - A deep learning model (U-Net) processes the images
to segment the foot outline, compute dimensional parameters (length, width,
arch height, instep height), and classify the arch type.
Page 14 of 30
[0090] Output Generation - The results are instantly displayed on-screen, including
the foot type classification, measurements, suggested shoe size, and optional
orthotic recommendations. The user can download the report or receive it via
email/QR code.
[0091] Optional Commercial Integration - In retail settings, the system may be linked
with a footwear database to recommend specific models or sizes based on the
user's foot profile.
[0092] This configuration ensures ease of use, minimal operator dependency, high
diagnostic accuracy, and fast throughput — ideal for high-traffic environments
like hospitals, clinics, shoe stores, schools, and wellness kiosks
[0093] This combination of AI-driven segmentation and dual-view camera calibration
in a kiosk format is novel, scalable, and highly adaptable.
[0094] In another preferred embodiment, the system comprises A foot placement
platform equipped with calibrated markers for accurate spatial reference;
Multiple high-resolution cameras strategically positioned to capture multiangle images of the plantar and medial aspects of the feet; A processing
module embedded with trained deep learning models for foot segmentation,
arch classification, and dimensional analysis; Measurement tools leverage a
dots-per-unit (DPU) calibration method to derive precise metrics such as foot
length, width, arch height, and instep height; A touch-enabled user interface
for seamless user interaction, real-time guidance, and visualization of
diagnostic results.
[0095] In another preferred embodiment, detection of various morphological
parameters
[0096] Foot Length and Foot Width detection (Top-View) - A top-view image of the
foot is captured using an overhead high-resolution camera, a U-Net-based
deep learning model segments the foot contour accurately; ArUco markers or
calibrated reference markers placed on the platform are used to determine
the millimeter-per-pixel (DPU) scaling factor in the top view. Using this DPU
calibration, the system calculates key dimensions- Foot length: Measured
from the heel to the tip of the longest toe; and Foot width: Measured across
the widest part of the forefoot. These real-world foot dimensions are stored
for further processing and visualization.
Page 15 of 30
[0097] Arch Detection and Height Estimation (Side View): A side-view image of the
foot is captured to analyze the foot's sagittal profile. Using the known foot
length (in mm) obtained from the top view, a scale-matching algorithm
computes the millimeter-per-pixel ratio for the side view by aligning the pixelbased foot length in the side image with the actual measured length. The UNet model segments the side foot contour. A baseline is drawn between the
heel and toe contact points. The maximum vertical distance from this baseline
to the arch contour is computed—this is the arch height. The arch height in
pixels is then converted to millimetres using the calibrated side-view scale.
a. Based on the computed arch height, the system classifies the foot
arch using the following table:
Category Arch Height
(mm)
Risk Level
High Arch > 20 mm Low
Normal Arch 15–20 mm Low
Low Arch 10–14 mm Moderate
Mild Flat Foot 5–9 mm High
Moderate
Flat Foot
1–4 mm Very High
Severe Flat Foot < 1 mm Extremely
High
[0098] Instep Height and Other Morphological Parameters: Instep height is computed
from the foot base to the highest point of the foot contour at 55% of the total
foot length, as this anatomically corresponds to the instep region.
[0099] Other biomechanical or morphological parameters can also be derived using
similar geometric rules and formulas by referencing specific foot regions (e.g.,
heel height, midfoot depth, etc.) along the foot length axis.
[0100] The present system provides following Output Parameters:
Page 16 of 30
[0101] Arch Height (mm): Vertical distance between the foot arch and baseline in
millimeters, computed from the side view using calibrated scaling.
[0102] Arch Type Classification: Automatically classified as High Arch, Normal Arch,
Low Arch, Mild Flat Foot, Moderate Flat Foot, or Severe Flat Foot based on
arch height.
[0103] Instep Height (mm): Height measured at 55% of foot length from heel,
providing insight into midfoot structure for orthotic and footwear
considerations.
[0104] Foot Dimensions such as Foot Length (mm): Distance from heel to longest
toe; and Foot Width (mm): Measured at the forefoot’s widest region.
[0105] Suggested Shoe Size: Estimated based on foot length and width using
standardized regional sizing systems (e.g., EU, UK, US).
[0106] Orthotic Recommendations (Optional): Based on arch type, instep height, and
pressure zones (if pressure mapping is integrated), the system may suggest
custom insole design parameters or recommend specific corrective footwear
types.
[0107] This invention offers a contactless, fast, and cost-effective solution for flat
foot detection and foot morphology analysis. It is suitable for deployment in
both clinical environments (for orthopedic diagnostics) and commercial
applications (such as custom footwear and orthotic insole fabrication). The
system bridges the gap between foot diagnostics and footwear customization,
eliminating the need for separate tools and specialized personnel.
[0108] In another preferred embodiment in Segmentation (image processing
methodology) the Foot contour segmentation is performed using a U-Netbased deep learning model, trained separately on plantar (top-view) and
lateral (side-view) foot image datasets. The model generates precise binary
masks outlining the foot structure, which serve as the basis for dimensional
and morphological analysis.
Page 17 of 30
[0109] In another preferred embodiment, the Foot Length and Width (Top View) is
computed from the segmented top-view mask, the system computes the
minimum-area enclosing rectangle around the foot contour. The resulting
rectangle yields four corners from which all side lengths are computed. The
longer edge corresponds to the foot length, and the shorter edge corresponds
to the foot width, both measured in pixels. Pixel-to-millimeter scaling
(PPU_top) is derived using ArUco markers or calibrated reference objects
placed on the platform:
PPU_top = marker_pixel_distance / marker_real_distance
[0110] Real-world dimensions is calculated using the following:
[0111] Foot Length (mm) = foot_length_pixels / PPU_top
[0112] Foot Width (mm) = foot_width_pixels / PPU_top
[0113] In another preferred embodiment, Arch Height (Side View) is computed as
follows
[0114] The lateral foot contour be denoted as:
y_b(x), where x ∈ [x_L, x_R]
[0115] A straight line y_t(x) = m·x + c is fitted to the foot base using least-squares
regression across the entire horizontal interval.
[0116] The arch height is calculated as the maximum vertical gap between this
baseline and the actual foot contour:
h_arch = max [ y_t(x) − y_b(x) ]
for x ∈ [x_L + 0.2·w, x_L + 0.8·w], where w = x_R − x_L
[0117] In another preferred embodiment, the Instep Height is computed:
[0118] The instep height is measured at approximately 55% of the total foot
length, which corresponds to the anatomical instep region.
[0119] Let x_instep = x_L + 0.55·w, then:
h_instep = | y_t(x_instep) − y_b(x_instep) |
[0120] In another preferred embodiment, Pixel-to-Unit Conversion (Side View) is
done in the following manner:
[0121] Since the actual foot length in millimeters is already known from the top
view, it is used to calibrate the pixel scale of the side view:
Page 18 of 30
[0122] Let L_max be the length of the foot in pixels in the side view.
[0123] Let L_top be the real-world foot length (mm) from the top view.
[0124] Then, PPU_side = L_max / L_top
[0125] Arch Height (mm) = (h_arch / PPU_side)
[0126] Instep Height (mm) = (h_instep / PPU_side)
[0127] In a preferred embodiment, the ArUco Tag Calibration is done where ArUco
tags are used to extract known reference lengths, Tag centroids and corners
are used to compute pixel-to-mm conversion factors. This ensures
consistency across hardware variations.
[0128] In a preferred embodiment, to provide Real-Life Image Results, the system
captures high-resolution images of the subject's foot from both the top view
(plantar) and the side view (lateral) using strategically positioned camera
modules. The example shown corresponds to the left foot, with the images
acquired simultaneously from the top and the left-side camera of the device.
Upon acquisition, the captured images are processed using the trained U-Net
segmentation model, which accurately delineates the foot boundary in each
view. The resulting predicted contours are overlaid on the original images for
visualization and verification purposes.
[0129] The combined results illustrate the system’s ability to provide real-time,
automated, and contactless estimation of biomechanical parameters such as
foot length and arch height, forming the basis for further diagnostic
interpretation and footwear customization.
Page 19 of 30
[0130] Test Results and System Accuracy
[0131] To evaluate the performance and reliability of the proposed self-operated
kiosk system, a series of controlled tests were conducted on a sample group
of users with varying foot anatomies. The system’s output was compared
against clinically verified measurements and expert orthopedic evaluations.
1. Flat Foot Detection Accuracy
[0132] The system demonstrated 100% accuracy in identifying flat feet (pes planus)
across all test cases.
[0133] The automated classification matched the expert orthopedic diagnosis in all
tested subjects (n = 50).
[0134] The U-Net-based segmentation model successfully detected subtle differences
in arch profiles, ensuring consistent and repeatable classification across
different foot types.
[0135] 2. Foot Length and Width Measurement Accuracy
[0136] Foot length and width were measured using the top-view image and ArUcocalibrated scaling.
[0137] The system’s results were compared to manual caliper-based measurements
performed by trained personnel.
[0138] The observed maximum deviation in foot length or width measurement was
within ±5 mm, which falls well within the acceptable tolerance range for both
Page 20 of 30
diagnostic and commercial footwear applications.
Parameter Mean Error (mm) Max Error (mm) Accuracy
(%)
Foot
Length
±2.8 mm ±5.0 mm 98.6%
Foot
Width
±3.2 mm ±5.0 mm 98.1%
Arch Type 0 (perfect match) 0 100%
[0139] 3. Processing Time -The complete analysis, including image capture,
segmentation, measurement, and output generation, was completed in under
60 seconds per subject.
[0140] 4. Repeatability and Reliability-Multiple runs on the same subject showed less
than 2% deviation in results, demonstrating excellent repeatability and
robustness of the system
[0141] The Present invention have multiple ecomonic benefits which are as follows:
[0142] Mass Deployment at Low Cost: Enables mass screening in hospitals, schools,
sports centers, and retail stores without the need for skilled operators.
[0143] Custom Insole Manufacturing: Facilitates low-cost, made-to-measure orthotic
insoles based on accurate, repeatable foot metrics.
[0144] Reduced Diagnosis Costs: Eliminates dependence on high-end medical
devices or professional services for initial assessment.
[0145] Commercial Upsell Opportunities: Retailers can use the kiosk to sell
personalized footwear, driving revenue through value-added services.
[0146] Data-Driven Insights: Aggregated data enables analytics, predictive foot
health models, and integration into digital health platforms.
[0147] Advantages of the present system over the existing systems:
1. Standalone Self-Operated Diagnostic Kiosk -This embodiment consists of a
complete, integrated unit comprising a foot placement platform, dual-view
imaging system (top and side cameras), a lighting module, a processing unit
embedded with deep learning algorithms, and a touchscreen display. It is
intended for deployment in clinical settings, orthotic centers, and retail footwear
Page 21 of 30
outlets. It enables users to conduct a full diagnostic cycle in less than 60 seconds
without operator assistance.
2. Portable Diagnostic Variant - In this embodiment, the invention is miniaturized
into a compact, mobile unit suitable for transport and field use. It includes a
detachable processing device (such as a tablet), battery operation, and
collapsible hardware. It is intended for rural healthcare camps, home-based
diagnostics, and outreach programs where portability is essential.
3. Retail Store Integration Embodiment - This version is embedded into a retail
footwear environment. The diagnostic data (foot length, width, arch type) is
integrated with the retailer’s sizing and inventory database to recommend
suitable footwear models. It may also be linked to customer loyalty programs or
in-store kiosks for personalized footwear suggestions.
4. Orthotic Manufacturing Interface - In this embodiment, the diagnostic output of
the system is formatted for direct compatibility with CAD or 3D printing
workflows. Foot geometry data can be exported in standard formats (e.g., STL,
OBJ) to facilitate the automated design and fabrication of custom orthopedic
insoles or footwear.
5. Enhanced Pressure and Gait Analysis Variant - A more advanced embodiment
may incorporate pressure sensors or force plates in the foot placement platform.
This allows real-time analysis of plantar pressure distribution and gait
characteristics, useful in physiotherapy, rehabilitation, and sports medicine
applications.
6. Cloud-Connected and Telemedicine Version - In this embodiment, the system is
equipped with internet connectivity for remote access and cloud-based storage.
Diagnostic data can be uploaded securely and accessed by healthcare
professionals for remote consultation, follow-up, and integration into electronic
medical records (EMR) systems.
7. Dual Foot Scanning System - This embodiment includes two-foot placement
platforms and an array of synchronized cameras, enabling simultaneous
scanning of both feet. This increases throughput and is suitable for institutional
settings such as schools, military screening, or corporate health checkups.
8. Voice-Guided and Multilingual System-For enhanced accessibility, this variation
includes multilingual voice instructions and large-format on-screen prompts,
Page 22 of 30
designed especially for elderly users or populations with low literacy. It ensures
intuitive use without technical assistance.
9. Report Generation and Digital Delivery Embodiment - In this embodiment, users
receive their diagnostic report in digital format via email, QR code, or app-based
notification. This supports eco-friendly operation, remote consultation, and
recordkeeping.
10. Integration with Public Health Systems - Future embodiments may include
integration with national or regional health information systems, allowing
aggregated foot health data to be used for epidemiological studies, policy
development, or insurance underwriting.
11.Contactless and Hygienic - While many existing solutions require users to place
their feet on contact-based sensors or mats, this system operates entirely
through cameras, reducing hygiene risks—especially important in post-pandemic
scenarios.
[0148] In a preferred embodiment, the system is integrated with 3D-printing systems
to generate custom insoles or corrective footwear based on patient-specific
foot scans.
[0149] In another preferred embodiment, the system when connected seamlessly
with cloud connectivity allow doctors to remotely review diagnostics and
prescribe treatments or custom footwear.
[0150] In another preferred embodiment, the system results provide real-time shoe
size recommendations and gait-type analysis can be integrated into footwear
e-commerce platforms or physical stores.
[0151] The present system may be deployed in mass scale in schools, sports
academies, and corporate wellness programs for early detection of
biomechanical issues.
[0152] The present system can also be used in long-term tracking of foot morphology
could be combined with activity data to detect degenerative foot disorders
over time.
Page 23 of 30
WE CLAIM:
1. A Novel Kiosk System [800] for Flat Foot Analysis and Foot Morphology using AI
Based Image Processing, the system comprising:
a. a foot placement platform [400] provided with stopper and boundaries, and
calibrated with Aruco marker for spatial accuracy;
b. high resolution camera modules [804,806] for capturing images;
c. a processing unit employing a U-Net-based deep learning model algorithm
to compute dimensional parameters of foot and classify foot arch types,
wherein dimensional parameters are medial arch height, instep height, foot
dimensions;
d. a display interface [100] controlled by the processing unit and integrated
with infrared touch screen panel for interactive user input and result
visualization; and
e. lighting units controlled by a control module comprising MOSFET based
LED controller to regulate LED operation based on system based
commands,
all positioned and operatively connected to a main body disposed with a
central mounting frame, curved side panels [600] on either side of central
mounting frame, background shield panel or side optical shield, a pair of
adjustable side arm unit or camera mount brackets [200] responsible for
holding cameras and lighting units, a display mount stand [100], cables and
connecting accessories.
2. The novel Kiosk System [800] for Flat Foot analysis and Foot Morphology using AI
Based Image Processing, as claimed in claim 1, wherein the foot placement
platform is disposed with a wide base plate of 550 mm with a distance of 120mm
between the outer edge of the foot placement area and the inner surface of the
side-arm unit, a 3D printed heel stopper at a distance of 60 mm from the rear
edge of the base to ensure proper foot placement during capturing of foot image,
and wherein the Aruco marker printed on vinyl is used for camera calibration and
spatial reference millimetre per pixel conversion.
Page 24 of 30
3. The novel Kiosk System [800] for Flat Foot analysis and Foot Morphology using AI
Based Image Processing, as claimed in claim 1, wherein the adjustable camera
mount brackets [200] on either side of the central mounting frame houses side–
facing cameras […] for taking side-view images and top cameras for taking topview images and lighting units, contributing to functional symmetry, and wherein
the adjustability of the camera mount brackets [200] enables precise alignment
of image sensors relative to the foot placement platform for optimal field of view
and focus.
4. The novel kiosk system [800] for flat foot analysis and foot morphology using AI
based Image processing, as claimed in claim 1,
wherein the display interface or unit [803] is of 1100mm height which
interacts with the user without straining or bending body;
wherein in the adjustable camera mount brackets [200], the camera
mounting slot of top-view capturing cameras are recessed by 19.33mm from
the front edge of the arm providing optimal field of view and camera
protection;
wherein the adjustable camera mount brackets [200], the camera mounting
slots of the side-view capturing cameras are disposed at the centre of the
vertical face [202] of the brackets, with the mounting slot positioned at
140.21 mm above the arm base or bracket base, and the side-view capturing
camera slots are positioned 37.60 mm apart horizontally from each other;
and the side-view camera are disposed at an inclination of 3.5 degree for
taking side view images of the mount providing optimal field of view and
accuracy in arch detection and side-profile analysis.
wherein the background shield panel or the side optical shield [602], each of
height 220 mm providing full vertical coverage to the lateral imaging filed;
panel width [603] of 558 mm ensure horizontal shield between foot and
background; depth or curvature extension [601] 84mm provides smooth
curved profile reducing internal reflections marinating optical clarity, and top
corners rounded with radius 40mm, blocks ambient background elements
Page 25 of 30
during image acquisition enhancing image contrast and segmentation,
provides soft user-safe edge;
wherein the lighting units [805…] are disposed in the upper horizontal portion
[202] of the adjustable camera mount brackets [200] at a distance of 46.02
mm from the edge of the arm and two lighting units are disposed at a distance
of 52.50 mm from one another, ensuring symmetrical light distribution across
imaging field, minimize shadows and provide image clarity during
segmentation and measurement;
5. The novel Kiosk System [800] for Flat Foot analysis and Foot Morphology using AI
Based Image Processing, as claimed in claim 1, wherein the lighting units are high
brightness LEDs, used for illumination or status indication during the scanning and
analysis process.
6. The novel Kiosk System [800] for Flat Foot analysis and Foot Morphology using AI
Based Image Processing, as claimed in claim 1, wherein display interface displays
foot profile including foot type classification, measurements, shoe size, orthotic
recommendations.
7. A method of flat foot analysis and foot morphology using AI Based Image Processing
through the system as claimed in claim 1, the method comprising:
a. positioning of user feet on the foot placement platform [400] properly
positioned using the 3D printed heel stopper and boundaries to ensure
proper foot placement during capturing of foot image, wherein the platform
is calibrated with the Aruco marker printed on vinyl to enable spatial
reference millimetre-per-pixel conversion and is used for camera calibration;
b. capturing of top-view image and side view of foot simultaneously,
wherein the top-view image is captured using a high resolution camera placed
in an overhead position above the foot placement platform, followed by foot
contour segmentation using U-net deep learning model algorithm,
determination of millimetre-per –pixel scaling factor using the ArUco markers
and calculation of foot length which is measured from heel to the tip of the
longest toe and foot width which is measured across the widest part of the
forefoot;
Page 26 of 30
wherein the side-view image of the foot is captured using high resolution
camera for analysing the sagittal profile of the foot, using the foot length
obtained from the top view, computation of millimetre-per-pixel ratio for the
side view using scale matching algorithm by aligning the pixel –based foot
length in the side image with the actual measured length, followed by side
foot contour segmentation using U-Net model; and
wherein the foot contour segmentation is performed using a U-Net based deep
learning model, trained separately on plantar (top-view) and lateral (sideview) foot image datasets, enabling the model to generate precise binary
masks outlining the foot structure forming the basis for dimensional and
morphological analysis;
c. classification of foot arch, first by computing the arch height, followed by
conversion of arch height from pixels into millimetre using calibrated side view
scale and identifying the foot arch based on a preloaded reference table,
wherein the arch height is calculated first by drawing a baseline between the
heel and toe contact points, then by calculating the maximum vertical distance
of the arch contour from this baseline;
d. calculation of instep height by computing the instep region from the foot base
to the highest point of the foot contour at 55% of the total foot length;
e. displaying in user-friendly interface the morphological parameters foot length
and width, arch height, arch type classification, Instep height, heel-to-toe,
and shoe size orthotic recommendations and completion of the process.
8. The method of flat foot analysis and foot morphology using AI Based Image
Processing through the system as claimed in claim 6, wherein the method completes
the analysis within 60 seconds and displays result in a user-friendly interface capable
of being downloaded and scanned through QR code.
9. The method of flat foot analysis and foot morphology using AI Based Image
Processing through the system as claimed in claim 6, wherein the result is influenced
by various morphological parameters obtained, in the manner,
Page 27 of 30
a. the foot length determines overall insole and shoe length, and to scale
anatomical positions including instep location at 55% of the foot length, and
is essential for matching size charts and toe box length;
b. the foot width defines the forefoot step and toe box dimensions and is
required for width-specific show manufacturing (narrow, standard, wide) to
prevent lateral compression;
c. the arch height is critical for identifying arch type (high, normal, low or flat)
which guides the contouring of medial longitudinal arch in the insole,
influences arch support height and stiffness in CAD modelling;
d. the instep height represents the height of the midfoot at 55% of the foot
length from the heel, defines upper volume and vamp height in shoes, affects
strap / closure positioning and pressure distribution over midfoot;
e. the arch height classification is computed based on arch height and threshold
values high arch (> 20mm), normal arch (15-20mm), low arch (10-14mm),
mild flatfoot (5-9 mm), severe flatfoot (< 1 mm); and
f. the Heel-to-Toe slope and volume estimation is calculated by comparing arch
height, heel height, instep height over length, and it defines insole transitions
from heel to forefoot.