DESC:
FORM 2

THE PATENTS ACT, 1970
(39 of 1970)
&
THE PATENT RULES, 2003

COMPLETE SPECIFICATION
(See Section 10 and Rule 13)

Title of invention:
METHOD AND SYSTEM FOR PLANOGRAM COMPLIANCE CHECK BASED ON VISUAL ANALYSIS

Applicant:
Tata Consultancy Services Limited
A company Incorporated in India under the Companies Act, 1956
Having address:
Nirmal Building, 9th floor,
Nariman point, Mumbai 400021,
Maharashtra, India

The following specification particularly describes the invention and the manner in which it is to be performed.
CROSS-REFERENCE TO RELATED APPLICATIONS AND PRIORITY

[001] The present application claims priority to Indian Provisional Patent Application No. 1927/MUM/2015, filed on May 16th, 2015, the entirety of which is hereby incorporated by reference.

FIELD OF THE INVENTION
[002] The embodiments herein generally relate to planogram compliance check, in particular for planogram compliance check based on visual analysis.

BACKGROUND OF THE INVENTION
[003] Planogram is a model that indicates the placement of retail products on shelves for customer ease. The research in retail marketing has shown that maintaining the planogram compliance can improve the overall profit by significant measure.

[004] Most of the previous approaches dealing with the tags or ancient accounting and inventory management methods. The RFID based approach faces the problem of exponential cost of sensor installation, and time-consuming hard work of attaching the tag with every product and removing them again at the billing counter. Additionally, in-spite of being capital intensive, desired results are not produced. The inventory based methods maintain the product log at the checkout point, and work with assumption that if a product is selected, it would be replaced at the same spot. A fallacy which is shared with some of the RFID/Sensor based approaches. These assumptions cannot be considered to be valid especially in conditions of high traffic or during the off-season when low stocks are maintained on the shelves. There have also been some image based analysis, which can cater to some of the problems discussed above. Nevertheless, some of these approaches rely on fixed or mounted cameras for monitoring requiring template images for all of the products available in the store. Also this approach is not able to properly test planogram compliance, if the front-face of product is not clearly visible.

SUMMARY OF THE INVENTION
[005] Before the present methods, systems, and hardware enablement are described, it is to be understood that this invention is not limited to the particular systems, and methodologies described, as there can be multiple possible embodiments of the present invention which are not expressly illustrated in the present disclosure. It is also to be understood that the terminology used in the description is for the purpose of describing the particular versions or embodiments only, and is not intended to limit the scope of the present invention which will be limited only by the appended claims.

[006] The present application provides a method and system for planogram compliance check based on visual analysis.

[007] In one aspect the present application discloses a computer implemented method for planogram compliance check based on visual analysis comprising steps of capturing at least one product shelf images using an image capture device (200). The method further comprises identifying a region of interest (ROI) from the at least one product shelf images using an image preprocessing module (210). The method further comprises generating at least one rectified shelf image by applying a perspective distortion rectification of the identified region of interest (ROI). In an embodiment, the perspective distortion rectification is based on projective geometry rectification. The disclosed method further comprises extracting one or more rows from the rectified shelf image wherein the rectified shelf image comprises a plurality of horizontal row partitions using a row extraction module (212). The method further may comprise applying Hausdroff distance based image map for occupancy estimation for each of the extracted one or more rows from the rectified shelf image based on a preexisting Hausdroff based image map generated for a predefined reference shelf image using an occupancy computation module (214). Finally the method may also comprise the step of deriving a product count and a placement information for each of the extracted one or more rows based on the extracted rows and the estimated occupancy using a product count and placement module (216).

[008] In another aspect the present application also discloses , the system (102) comprising an image capture device (200) operatively coupled to the system, a processor (202) and an interface (204), a memory (206) coupled to said processor (202), the system (102) comprising, the image capture device (200) configured to capture at least one product shelf images, an image preprocessing module (210) configured to identify a region of interest (ROI) from the at least one product shelf images. Further the image preprocessing module (210) is configured generate at least one rectified shelf image by applying a perspective distortion rectification of the identified region of interest (ROI). In an embodiment the perspective distortion rectification is based on projective geometry rectification. The system (102) further comprises a row extraction module (212) configured to extract one or more rows from the rectified shelf image wherein the rectified shelf image comprises a plurality of horizontal row partitions, an occupancy computation module (214) configured applying Hausdroff distance based image map for occupancy estimation for each of the extracted one or more rows from the rectified shelf image based on a preexisting Hausdroff based image map generated for a predefined reference shelf image and a product count and placement module (216) configured to derive a product count and placement information for each of the extracted one or more rows based on the extracted rows and the estimated occupancy.

BRIEF DESCRIPTION OF THE DRAWINGS
[009] The foregoing summary, as well as the following detailed description of preferred embodiments, are better understood when read in conjunction with the appended drawings. For the purpose of illustrating the invention, there is shown in the drawings exemplary constructions of the invention; however, the invention is not limited to the specific methods and system disclosed. In the drawings:

[0010] Figure 1: a network implementation of a system for planogram compliance check based on visual analysis, in accordance with an embodiment of the present subject matter;

[0011] Figure 2: illustrates the system for planogram compliance check based on visual analysis, in accordance with an embodiment of the present subject matter;

[0012] Figure 3: shows a flow chart illustrating a method for planogram compliance using visual analysis of different products in accordance with an example embodiment of the present disclosure;

[0013] Figure 4: shows a flow chart illustrating a method for planogram compliance using visual analysis for Partial or Complete presence of different products in accordance with an example embodiment of the present disclosure;

[0014] Figure 5: shows a flow chart illustrating a Self-Hausdroff map creation in accordance with an example embodiment of the present disclosure; and

[0015] Figure 6: shows a flow chart illustrating a method of row extraction in accordance with an example embodiment of the present disclosure.

DETAILED DESCRIPTION OF THE INVENTION
[0016] Some embodiments of this invention, illustrating all its features, will now be discussed in detail.

[0017] The words "comprising," "having," "containing," and "including," and other forms thereof, are intended to be equivalent in meaning and be open ended in that an item or items following any one of these words is not meant to be an exhaustive listing of such item or items, or meant to be limited to only the listed item or items.

[0018] It must also be noted that as used herein and in the appended claims, the singular forms "a," "an," and "the" include plural references unless the context clearly dictates otherwise. Although any systems and methods similar or equivalent to those described herein can be used in the practice or testing of embodiments of the present invention, the preferred, systems and methods are now described.

[0019] The disclosed embodiments are merely exemplary of the invention, which may be embodied in various forms.

[0020] The elements illustrated in the Figures inter-operate as explained in more detail below. Before setting forth the detailed explanation, however, it is noted that all of the discussion below, regardless of the particular implementation being described, is exemplary in nature, rather than limiting. For example, although selected aspects, features, or components of the implementations are depicted as being stored in memories, all or part of the systems and methods consistent with the attrition warning system and method may be stored on, distributed across, or read from other machine-readable media.

[0021] The techniques described above may be implemented in one or more computer programs executing on (or executable by) a programmable computer including any combination of any number of the following: a processor, a storage medium readable and/or writable by the processor (including, for example, volatile and non-volatile memory and/or storage elements), plurality of input units, and plurality of output devices. Program code may be applied to input entered using any of the plurality of input units to perform the functions described and to generate an output displayed upon any of the plurality of output devices.

[0022] Each computer program within the scope of the claims below may be implemented in any programming language, such as assembly language, machine language, a high-level procedural programming language, or an object-oriented programming language. The programming language may, for example, be a compiled or interpreted programming language. Each such computer program may be implemented in a computer program product tangibly embodied in a machine-readable storage device for execution by a computer processor.

[0023] Method steps of the invention may be performed by one or more computer processors executing a program tangibly embodied on a computer-readable medium to perform functions of the invention by operating on input and generating output. Suitable processors include, by way of example, both general and special purpose microprocessors. Generally, the processor receives (reads) instructions and data from a memory (such as a read-only memory and/or a random access memory) and writes (stores) instructions and data to the memory. Storage devices suitable for tangibly embodying computer program instructions and data include, for example, all forms of non-volatile memory, such as semiconductor memory devices, including EPROM, EEPROM, and flash memory devices; magnetic disks such as internal hard disks and removable disks; magneto-optical disks; and CD-ROMs. Any of the foregoing may be supplemented by, or incorporated in, specially-designed ASICs (application-specific integrated circuits) or FPGAs (Field-Programmable Gate Arrays). A computer can generally also receive (read) programs and data from, and write (store) programs and data to, a non-transitory computer-readable storage medium such as an internal disk (not shown) or a removable disk.

[0024] Any data disclosed herein may be implemented, for example, in one or more data structures tangibly stored on a non-transitory computer-readable medium. Embodiments of the invention may store such data in such data structure(s) and read such data from such data structure(s).

[0025] The present application provides a computer implemented method and system for visual analysis for planogram compliance. The disclosed subject matter provides a system (200) comprising an image capture device communicatively coupled with a system comprising an image preprocessing module (210), a row extraction module (212), an occupancy detection module (214) and a count and placement identification module (216).

[0026] It should be noted that the description merely illustrates the principles of the present subject matter. It will thus be appreciated that those skilled in the art will be able to devise various arrangements that, although not explicitly described herein, embody the principles of the present subject matter and are included within its spirit and scope. Furthermore, all examples recited herein are principally intended expressly to be only for explanatory purposes to aid the reader in understanding the principles of the invention and the concepts contributed by the inventor(s) to furthering the art, and are to be construed as being without limitation to such specifically recited examples and conditions.

[0027] Referring now to Fig 1, a network implementation 100 of a system 102 for planogram compliance using visual analysis is illustrated, in accordance with an embodiment of the present subject matter. Although the present subject matter is explained considering that the system 102 is implemented on a server, it may be understood that the system 102 may also be implemented in a variety of computing systems, such as a laptop computer, a desktop computer, a notebook, a workstation, a mainframe computer, a server, a network server, and the like. In one implementation, the system 102 may be implemented in a cloud-based environment. It will be understood that the system 102 may be accessed by multiple users through one or more user devices 104-1, 104-2…104-N, collectively referred to as user devices 104 hereinafter, or applications residing on the user devices 104. Examples of the user devices 104 may include, but are not limited to, a portable computer, a personal digital assistant, a handheld device, and a workstation. The user devices 104 are communicatively coupled to the system 102 through a network 106.

[0028] In one implementation, the network 106 may be a wireless network, a wired network or a combination thereof. The network 106 can be implemented as one of the different types of networks, such as intranet, local area network (LAN), wide area network (WAN), the internet, and the like. The network 106 may either be a dedicated network or a shared network. The shared network represents an association of the different types of networks that use a variety of protocols, for example, Hypertext Transfer Protocol (HTTP), Transmission Control Protocol/Internet Protocol (TCP/IP), Wireless Application Protocol (WAP), and the like, to communicate with one another. Further the network 106 may include a variety of network devices, including routers, bridges, servers, computing devices, storage devices, and the like.

[0029] Referring now to Fig. 2, a system (102) for planogram compliance check based on visual analysis is illustrated according to an embodiment of the disclosed subject matter

[0030] According to one embodiment, referring to Fig. 2, the system (102) is configured for planogram compliance check based on visual analysis more specifically, the system (102) comprises an image capture device (200) operatively coupled to the system, a processor (202) and an interface (204), a memory (206) coupled to said processor (202). Further more specifically, in an embodiment of the disclosed subject matter an image capture device (200) configured to capture at least one product shelf images. In another embodiment the system (102) further comprises an image preprocessing module (210) further configured to identify a region of interest (ROI) from the at least one product shelf images. In another embodiment, the image preprocessing module (210) is further configured to generating at least one rectified shelf image by applying a perspective distortion rectification of the identified region of interest (ROI). In an aspect the perspective distortion rectification is based on projective geometry rectification.

[0031] Capturing the product shelf image, wherein a perspective distortion rectification of images are captured, by any image capturing device available in the art, selecting the corner points of the product shelf using the touch screen of the image capturing device, rectifying perspective distortion in the image by using the corner points by applying conventional projective geometry rectification.

[0032] Further, referring again to Fig. 2 the system (102) further comprises a row extraction module (212) which may be configured to extract one or more rows from the rectified shelf image. In an aspect the rectified shelf image comprises a plurality of horizontal row partitions, wherein row extraction is based on identification of these horizontal row partitions on the rectified shelf image. In another embodiment the row extraction may also be applied to a preexisting reference shelf image. In yet another embodiment row extraction comprises the steps of generating a chg-img by finding all vertical changes in the image using Sobel derivative, applying Hough transform on the chg-img to detect all possible horizontal lines; and extracting the one or more rows by identifying most prominent horizontal lines in the shelf. The details of the row extraction process are explained further below in this application with reference to Fig. 6.

[0033] Referring again to Fig. 2, the system (102) further comprises an occupancy computation module (214) configured applying Hausdroff distance based image map for occupancy estimation for each of the extracted one or more rows from the rectified shelf image. In an embodiment the occupancy estimation module (214) is further configured to apply Hausdroff based image map such that occupancy is estimated as one of partially missing and completely missing, wherein partially missing refers to the situation where the a product is partially missing while completely missing is selected when a product is completely missing. In an embodiment the occupancy detection is based on matching the Hausdroff based image map for the rectified shelf image with a preexisting Hausdroff based image map for a predefined reference shelf image.

[0034] Referring again to Fig.2 the system (102) further comprises a product count and placement module (216) which configured to derive a product count and placement information for each of the extracted one or more rows based on the extracted rows and the estimated occupancy. In an embodiment the product count and placement module (216) is configured to detect a product based on a texture properties of a product image wherein the product image is a part of the rectified product shelf image, by matching the texture properties of the product image with a predefined reference product shelf image. Further product count and placement module (216) may be configure to eliminate one or more false positives product detection by matching the color histogram of the product image with the predefined reference product shelf image.

[0035] Referring now to Fig. 3, a flowchart a flow chart illustrating a method for planogram compliance using visual analysis of different products in accordance with an example embodiment of the present disclosure.

[0036] The process starts at the step 302 at least one product shelf image is captured using an image captured device (200) wherein the image capture device is operatively coupled to the system.

[0037] At the step 304 and 306 preprocessing is performed on the image wherein at the step 304, a region of interest (ROI) is identified from the captured at least one product shelf images. At the step 306 at least one rectified shelf image is generated by applying a perspective distortion rectification of the identified region of interest (ROI). In an aspect the perspective distortion rectification is based on projective geometry rectification.

[0038] At the step 308, one or more rows are extracted from the rectified shelf image. In an aspect the rectified shelf image comprises a plurality of horizontal row partitions, wherein row extraction is based on identification of these horizontal row partitions on the rectified shelf image. In another embodiment the row extraction may also be applied to a preexisting reference shelf image. In yet another embodiment row extraction comprises the steps of generating a chg-img by finding all vertical changes in the image using Sobel derivative, applying Hough transform on the chg-img to detect all possible horizontal lines; and extracting the one or more rows by identifying most prominent horizontal lines in the shelf.

[0039] At the step 310 Hausdroff distance based image map is applied for occupancy estimation for each of the extracted one or more rows from the rectified shelf image. In an embodiment Hausdroff based image map may be applied such that occupancy is estimated as one of partially missing and completely missing, wherein partially missing refers to the situation where the a product is partially missing while completely missing is the situation when a product is completely missing. Occupancy estimation results may be for each of the extracted one or more rows or for the entire shelf. In an embodiment the occupancy detection is based on matching the Hausdroff based image map for the rectified shelf image with a preexisting Hausdroff based image map for a predefined reference shelf image.

[0040] At the step 312 a product count and placement information is derived for each of the extracted one or more rows based on the extracted rows and the estimated occupancy. In an embodiment a product is detected based on a texture properties of a product image wherein the product image is a part of the rectified product shelf image, by matching the texture properties of the product image with a predefined reference product shelf image. Further in another embodiment one or more false positives product detection may be eliminated by matching the color histogram of the product image with the predefined reference product shelf image.

[0041] The above described method is now explained in detail with reference to the drawings FIG. 4 to FIG. 6

[0042] Computing occupancy based on Hausdorff map - calculating a local distance map for product shelf image captured and a reference image based on Hausdorff metric and comparing the product shelf image and the referenced image using the below equation:
Equation 1: H(A;B) = max{h(A;B); h(B;A)}
Equation 2: h(A;B) = max (a€2) {min (b€B) {d(a; b)}}

[0043] Here A, and B are the local areas of reference image and product shelf image Whereas a and b in equation 2 are smaller divisions in A and B, i.e., A and B are the super-masks, and a and b denote sub-masks in corresponding super-masks. h(A;B) in the equation 1 denotes the directed Hausdorff distance as described in equation 2.

[0044] Computing occupancy method using Hausdorff metric based distance map. It includes individual identification of complete and partial missing cases.

Identification of completely missing case
[0045] In reference to Fig. 4, for identification of completely missing product, Euclidean distance is applied as the comparison metric d (a; b) with all computations performed in RGB color space. The metric and the selected color space identifies the missing, misplaced and wrongly placed products using the marked difference in the RGB intensity values with the reference image. The sub-masks a and b are expressed as the mean value of pixel intensities in RBG planes in their defined region, using these values a, b, and d as Euclidean distance, the product shelf distance map is plotted and Otsu's thresholding criterion is applied.

[0046] In reference to FIG. 4, for identification of partially missing product, Hausdroff map based on Chi-Square distance as shown in equation 3
Equation 3: d(H1;H2) = ? (H1(I) - H2(I))2 /H1(I)

I
[0047] Referring to equation 3, H1 and H2 are the histograms of the sub-masks of reference and product shelf image. The noisy artifacts generated during this process is addressed by generating self Hausdorff map of the reference image using horizontal and vertical translation as shown in the FIG. 5.
Product shelf row identification
[0048] In reference to FIG. 4, the steps include deriving all the vertical changes in the image using Sobel derivative with an output chg_img, followed by applying Hough transform on the chg_img to detect all possible straight lines. An automated solution is provided for selection of related parameters, i.e., Sobel derivative threshold, and Voting threshold in Hough transform based line detection. The solution includes an algorithm as follows:
Line Reduction Algorithm
1: procedure reduceLines (sobT h; hoghTh; img)
2: maxLines <--- (img.ht/10)
3: while linDtcd > maxLines do
4: linDtcd <---- HoughLines (sobTh, hoghTh, img)
5: incr <----- (linDtcd/maxLines)
6: hoghTh <----- (hoghTh + incr)
7: sobTh <----- sobTh + (incr/2)
8: end while
9: return sobT h; hoghTh
10: end procedure

Product Count and Placement Computation
[0049] A row level product count from the above steps assumes the availability of exact product templates. The product count and computation follows two steps: a) detecting product based on texture properties of the product image, b) Eliminating false positives using color features.

[0050] a) Detecting product based on texture properties of the product image: applying a SURF feature based detection for first level of product detection. Identifying slide window with respect to the aspect-ratio of the product-template, wherein the height of the window being equal to the height of the.

[0051] b) Eliminating false positives using color features: applying second level of product detection for removal of false positives by matching the color histogram between the template image and the detected regions in example image.

[0052] Further, in another embodiment the method includes the step of applying normalized histogram and using Bhattacharya distance for measuring the similarity between two distributions. The Bhattacharya distance between two equalized histograms h1, and h2 is defined as in eq. 4.

Equation 4: d(h1; h2) = v(1- ? (v(h1(i) *h2(i)))
i

[0053] The above described system performs the operation in accordance to the one or more embodiments disclosed in the present disclosure from FIG. 4 to 6, however it may be apparent to a person skilled in the art that the system and method disclosed in this application may be implemented using a plurality of embodiments and these embodiments may only be construed for understanding the working of the present system and method and not limit the scope of the instant application.
,CLAIMS:1. A method for planogram compliance check based on visual analysis comprising processor implemented steps of:
capturing at least one product shelf images using an image capture device (200);
identifying a region of interest (ROI) from the at least one product shelf images using an image preprocessing module (210);
generating at least one rectified shelf image by applying a perspective distortion rectification of the identified region of interest (ROI) by the image preprocessing module (210);
extracting one or more rows from the rectified shelf image wherein the rectified shelf image comprises a plurality of horizontal row partitions using a row extraction module (212);
applying Hausdroff distance based image map for occupancy estimation for each of the extracted one or more rows from the rectified shelf image based on a preexisting Hausdroff based image map generated for a predefined reference shelf image using an occupancy computation module (214);
deriving a product count and a placement information for each of the extracted one or more rows based on the extracted rows and the estimated occupancy using a product count and placement module (216).

2. The method according to claim 1 wherein applying Hausdroff based image map further comprises estimating occupancy as one of partially missing and completely missing by matching the Hausdroff based image map for the rectified shelf image with the preexisting Hausdroff based image map generated for a predefined reference shelf image.

3. The method according to claim 1 wherein extracting one or more rows comprises:
generating a chg-img by finding all vertical changes in the image using Sobel derivative;
applying Hough transform on the chg-img to detect all possible horizontal lines; and
extracting the one or more rows by identifying most prominent horizontal lines in the shelf.

4. The method according to claim 1 wherein the deriving a product count and placement information further comprises:
detecting a product based on a texture properties of a product image wherein the product image is a part of the at least one product shelf image, by matching the texture properties of the product image with a predefined reference product shelf image; and
eliminating one or more false positives product detection by matching the color histogram of the product image with the predefined reference product shelf image.

5. A system (102) for planogram compliance check based on visual analysis, the system (102) comprising an image capture device (200) operatively coupled to the system, a processor (202) and an interface (204), a memory (206) coupled to said processor (202), the system (102) comprising:
the image capture device (200) configured to capture at least one product shelf images;
an image preprocessing module (210) configured to identify a region of interest (ROI) from the at least one product shelf images;
the image preprocessing module (210) further configured to generate at least one rectified shelf image by applying a perspective distortion rectification of the identified region of interest (ROI);
a row extraction module (212) configured to extract one or more rows from the rectified shelf image wherein the rectified shelf image comprises a plurality of horizontal row partitions;
an occupancy computation module (214) configured applying Hausdroff distance based image map for occupancy estimation for each of the extracted one or more rows from the rectified shelf image based on a preexisting Hausdroff based image map generated for a predefined reference shelf image; and
a product count and placement module (216) configured to derive a product count and placement information for each of the extracted one or more rows based on the extracted rows and the estimated occupancy.
6. The system according to claim 5 wherein the occupancy computation module () is further configured to estimate occupancy as one of partially missing and completely missing by matching the Hausdroff based image map for the rectified shelf image with a preexisting Hausdroff based image map generated for a predefined reference shelf image.

7. The system according to claim 5 the row extraction module (212) is configured to:
generate a chg-img by finding all vertical changes in the image using Sobel derivative;
apply Hough transform on the chg-img to detect all possible horizontal lines; and
extract the one or more rows by identifying most prominent horizontal lines in the shelf.

8. The system according to claim 5 wherein the product count and placement module (216) is further configured to
detect a product based on a texture properties of a product image wherein the product image is a part of the rectified product shelf image, by matching the texture properties of the product image with a predefined reference product shelf image ; and
eliminate one or more false positives product detection by matching the color histogram of the product image with the predefined reference product shelf image.