CLIAMS: A concentric ring array marker for landing of an unmanned aerial vehicle (UAV), wherein the marker comprises a variable width concentric ring pattern;
wherein the width of an outermost ring of the concentric ring marker is widest with successive inner rings having progressively narrower widths, wherein the concentric ring widths have a linearly increasing relationship from the centre towards the circumference of the concentric ring marker.
The concentric ring marker as claimed in claim 1, wherein the width of said outermost projected ring of the concentric ring marker at various altitudes remains same.

The concentric ring marker as claimed in claim 1, wherein the width of the outermost ring of the concentric ring marker in two successive captured images remains same as provided below:
width of the outermost ring projection =
f/D_1 [D_1× tan??+d_1/2]-f/D_1 [D_1× tan??-d_1/2] = (f·d_1)/D_1 = (f·d_2)/D_2
where d1 and d2 are width of two successive rings in the concentric ring marker respectively, having their distance from a down-looking camera being D1 and D2, and f is the focal length of the down-looking camera.

The concentric ring marker as claimed in claim 1, wherein the widths of the successive rings in the concentric ring marker forms an arithmetic progression. ,TagSPECI:TECHNICAL FIELD
The present disclosure relates, in general, to vision-based landing of unmanned aerial vehicles (UAVs), and in particular, to a method of concentric ring array marker based landing of UAVs.
BACKGROUND
Conventionally, usage of Unmanned Aerial Vehicle (UAV) for various applications is known in the art. These applications involve remote sensing typically in outdoor areas such as urban areas and remote rural areas. For example, the urban area applications include UAV usage by fire brigades, police or disaster control and emergency response teams. Similarly, the remote rural area applications include the UAV usage for periodic monitoring of linear infrastructures, such as power lines, oil/gas pipelines and the like.
Traditionally, vision-based landing systems have been proposed in relation with landing of a UAV on a precisely located target. Further, by estimation of 3D relative positioning and pose estimation using monocular vision techniques, the landing errors have been attempted to be reduced. Further, to aid such estimation, usage of a marker-based tracking on a landing platform or a target, due to its lesser computational complexity, has been suggested. Use of concentric features such as concentric circular rings and squares has also been suggested. The marker-based landing platform includes a concentric ring array as the landing pattern printed on the target location. Additionally, the symmetry of these shapes can be easily detected and they convey a reliable scale.
However, one drawback of the aforesaid imaging techniques is partial imaging. The partial imaging problem also throws up another issue of too thin or too thick projection of rings on the camera mounted on the UAV, as and when the UAV descends towards its target. More precisely, for a specific ring on the marker, the projection becomes successively thicker as the UAV loses altitude.
Further, another challenge regarding the varying width of a ring’s projection is that at high altitudes in real scenarios, e.g. from where the UAV controller moves into landing state, projections of the ring will be too thin. This in turn impacts the outer-to-inner ring radius ratio, which is used to track/detect the rings and hence the landing marker pattern on the target. This problem arises irrespective of the landing marker pattern being partially imaged or fully imaged.
To overcome the problem of partial imaging such as varying width of a ring’s projection at various altitudes, use of variable-width markers for landing of a UAV has been suggested. If variable-width marker based targets are designed for landing of the UAV, it is possible to detect these targets against varying ambient light conditions. In a suggested variable-width marker, a series of concentric circles with exponentially distributed radii is formed. It has also been suggested that, these variables with markers can be detected using the Hough Transform (HT) technique.
However, the conventional variable-width markers do not provide, explicitly, a relationship between any of two variable width rings that can be used to engineer the marker.
Therefore, there is a need for a system that limits the aforementioned drawbacks in concentric ring marker based landing of an unmanned aerial vehicle.
OBJECTS
An object of the present disclosure is to provide a concentric ring array marker based landing platform for an unmanned aerial vehicle (UAV).
DEFINITIONS OF TERMS USED IN THE COMPLETE SPECIFICATION
The expression “waypoint” used hereinafter in this specification refers to a position chosen as a destination for navigation of a route. A route may have one or more waypoints. That is, a route is composed of waypoints, including at least one final waypoint, and one or more intermediate waypoints.
The expression “position” used hereinafter in this specification refers to a location in the air or over the ground. ‘Position’ is typically specified as Earth coordinates, latitude and longitude. A specification of position may also include altitude.
SUMMARY
This summary is provided to introduce concepts related to a concentric ring array marker based landing platform for unmanned aerial vehicles (UAVs), which is further described below in the detailed description. This summary is neither intended to identify essential features of the present disclosure nor is it intended for use in determining or limiting the scope of the present disclosure.
The present disclosure provides a concentric ring array marker based navigation system for landing of a UAV. The concentric ring array marker has a variable width concentric ring pattern. In the variable width concentric ring pattern applied in this disclosure, the width of an outermost ring of the concentric ring array marker is widest with successive inner rings have progressively narrower width. In accordance with the preferred embodiment of this disclosure, the concentric ring widths have a linearly increasing relationship from the centre to the circumference of the concentric ring marker.
BRIEF DESCRIPTION OF THE ACCOMPANYING DRAWINGS
The detailed description is described with reference to the accompanying figures. In the figures, the left-most digit(s) of a reference number identifies the figure in which the reference number first appears. The same numbers are used throughout the drawings to reference like features and modules.
Fig. 1A illustrates a conventional concentric marker array defining a landing pattern, suggested in the prior art.
Fig. 1B illustrates the partial imaging problem which the array of Fig. 1A suffers.
Fig. 2A illustrates a suggested pattern of concentric ring marker with variable width concentric rings, according to an implementation of the present disclosure.
Fig. 2B illustrates a contrasting graphical representation of the concentric ring width having a linear increase relationship (Fig. 1B) against an exponential increase relationship (Fig. 1A).
Fig. 3 illustrates an exemplary constant projection of two successive rings, according to an implementation of the present disclosure.
Fig. 4 illustrates an exemplary constant projection for the outermost ring imaged at two successive time instants during the descend, in accordance with an embodiment of the present disclosure.
DETAILED DESCRIPTION
The present disclosure relates to a concentric ring array marker based navigation system for landing of an unmanned aerial vehicle.
The present system is not limited to the specific embodiments described herein. In addition, modules of each system can be practiced independently and separately from other modules described herein. Each module can be used in combination with other modules.
Many photogrammetry applications make use of targets that are placed in an observed scene and enable very precise measurements of the position of a point in the different available views. In order to achieve greater automation, some targets may contain unique information or pattern so that they may be uniquely identified.
The present disclosure provides a concentric ring array marker based target that is uniquely identified by an UAV as an intended landing location. The concentric ring array marker is designed in such a way that the width of an outermost ring is widest, while successive inner rings have a progressively narrower width. Further, the width ratio of two successive rings must follow a linear relationship with the ratio of the radii. Thus, if ‘y’ is the width of a ring, ‘y’ is related to the radii of the adjacent ring as a function of ‘x’ which is the radii of the previous adjacent ring by the equation:
y=mx + c
(where m stands for a functional relationship and generally a gradient of a straight line passing through y, y1, y2….. yn, which are the widths of successive rings in the marker, and c is the radius of innermost, 0th ring)
Further, the present disclosure provides a concentric ring array marker based navigation system for landing of a UAV. The disclosed system includes a target, of predetermined geometry, positioned at one or more intended landing locations. The system further includes one or more sensors coupled to each UAV, such that at least one sensor such as a ground looking camera is aligned with the direction of movement of the UAV, and captures one or more images in a waypoint in the flight path of and/or direction of movement of the UAV. The system further includes at least one processor-based device such as a flight operation module, which estimates an altitude of the UAV based on the identified target through a plurality of sensors such as 6/9-DOF. To this end, the system estimates a 3-D position of the UAV relative to the identified target based on fusion of signals from different GPS, altimeter and said identified target. Further, the system directs the UAV to start descending towards the target positioned at an intended landing location based on the estimated altitude and the estimated 3-D position of the UAV.
The present disclosure provides a concentric ring marker having a unique pattern applied or formed thereon. The width of an outermost ring of the concentric ring marker is widest with successive inner rings having progressively narrower width. The concentric ring widths have a linearly increasing relationship from the centre to the circumference of the concentric ring marker. Thus, due to the partial imaging problem, at a certain altitude, only some of the rings, typically, the inner rings may get projected on the ground-looking camera such as CCD/CMOS camera mounted on the UAV. Unlike the exponential relationship for widths suggested in the prior art wherein there is approximately 70% shrinkage in detection, there is a little resemblance of sameness which results in a similar looking pattern while landing from different heights. There are distinct variable width rings in the marker of the present disclosure whereas in the exponentially increasing width markers the inner rings being thinner, the pattern may not be imaged at all below a certain landing height.
Therefore, the present disclosure provides a variable width concentric ring array marker having a unique ring width relationship. The ring width relationship provides that the outermost ring in the projected image, at two successive capture moments, has the same width. Though, the various rings in the concentric ring array marker provide that the rings in the actual concentric ring array marker are sufficiently placed apart, as the UAV descends, successive rings become the outermost rings in the projected image. In another implementation, the width of the successive rings in the concentric ring marker forms an arithmetic progression. In another implementation, the concentric ring widths have a linearly increasing relationship from the centre to the circumference of the concentric ring marker.
These and other advantages of the present subject matter would be described in greater detail in conjunction with the following figures. While aspects of described systems and methods of concentric ring array marker based landing of an unmanned aerial vehicle may be implemented in any number of different computing systems, environments, and/or configurations, the embodiments are described in the context of the following exemplary system(s).
Fig. 1A illustrates a conventional concentric ring array as the landing pattern 100. The conventional concentric ring array 100 includes a plurality of concentric rings with equal width. For an example, one array of the conventional concentric ring array 100 has four concentric white rings. The outer ring has a ratio of inner-to-outer radius is 85%, while next successive rings towards the center has this ratio as 75%, 65% and 50%. Therefore, there is no clear methodology provided by the conventional arts as to how these ratios have been arrived at.
Fig. 1B illustrates the partial imaging problem where the projection of rings on the camera 111 are very thin or very thick. The partial imaging problem throws the issue of too thin or too thick projection of rings on the ground-looking camera such as CCD/CMOS camera mounted on the UAV, as the UAV descends towards its target mark. More specifically, for a specific ring on the marker, the projection becomes successively thicker as the UAV loses altitude. Therefore, the problem with varying width of a ring’s projection is that at high altitudes in real scenarios, for example from where the UAV moves into landing state (may be 50-60 meters), projections of the ring may be too thin. This in turn impacts the outer-to-inner ring radius ratio, which is used to track/detect the rings and hence the landing marker. This problem happens irrespective of the pattern being partially imaged or fully imaged.
Fig. 2A illustrates a pattern 200 for a concentric ring array marker with variable width concentric rings, according to an implementation of the present disclosure. The concentric ring marker 200 with variable width concentric rings provides that the width of an outermost ring is widest, with successive inner rings have progressively narrower width. Therefore, the concentric ring width has a linear increasing relationship from the centre to the circumference of the concentric ring marker. Thus, due to the partial imaging problem, at a certain altitude, only some of the rings, particularly, inner rings may get projected on the ground-looking camera such as CCD/CMOS camera mounted on the UAV. The advantage of the pattern of the marker 200 is that the width of the outermost projected ring at various altitudes remains the same. Therefore, the projected part of the concentric ring marker 200 is similar across the multiple projections captured in a sequence of camera captures.
Fig. 2B illustrates a graphical representation of the concentric ring width having a linearly increasing relationship against an exponentially increasing relationship. The graphical representation suggests that the linearly increasing relationship can be represented based on the equation y= mx+c, where ‘y’ is a width of the concentric ring, ‘x’ is an integer representing the ring in a sequence starting from the centre, ‘c’ is a constant, and ‘m’ being the gradient of the line joining the integers from the centre to circumference.
Further, the plot, as provided in Fig. 2B, illustrates that in the linear relationship the width of the next ring is halved, whereas the width of the next ring in an exponentially related sequence goes down by 1/7th. This implies that at the next imaging epoch during landing, the outermost ring size may be 2/7th of the outermost ring size detected in the previous epoch. This amount to approximately 70% shrinkage in detection, and at some point, the outermost ring width may become small enough not to get imaged even in 1-pixel resolution from a particular camera/UAV altitude. The inner rings being even thinner, the pattern may not be imaged at all.
Fig. 3 illustrates exemplary projections of two successive rings, according to an implementation of the present disclosure. It must be noted that the frame rate of capturing a sequence of images by a camera mounted on the UAV is, generally fixed. Assuming, the inter-frame-capture interval to be “T”. In this period of “T” seconds, assuming that the descending velocity is constant, the UAV is expected to move “s” units between the capture of two successive frames. The almost-constant, closed-loop-controlled velocity is a fair assumption. Further, the camera field of view may be denoted by 2*?, and camera focal length may be “f”, both being constants given a specific camera model mounted on the UAV. An exemplary model for capture of successive frames is depicted in Fig. 3. Further, it may also be assumed that so as not to miss out on the imaging of intermediate rings between successive captures, the separation between rings, i.e the inter-ring distance, as observed in the “elevation” view, cannot be less than s*tan(?).
Fig. 4 illustrates an exemplary projection 400 for the outermost ring, in accordance with an embodiment of the present disclosure. The ring width relationship in accordance with this disclosure provides that the outermost ring in the projected image, for two successive captures has the same width. Though, the various rings in the concentric ring marker provides that the rings in the actual concentric ring marker are sufficiently placed apart, as the UAV descends, successive rings become the outermost rings in the projected image. Assuming two such successive rings in the marker to be of width d1 and d2 respectively, having their distance from the camera pinhole (pinhole to ring center) being D1 and D2 respectively, the width of outer ring projection = f/D_1 [D_1× tan??+d_1/2]-f/D_1 [D_1× tan??-d_1/2]
= (f·d_1)/D_1
Similarly, width of outer ring projection = (f·d_2)/D_2

If the ring projection are to be same, the d_1/D_1 =d_2/D_2
However, D2 = D1 – s , where s is the constant distance covered by UAV between successive frame captures.
?(= ) d_1/D_1 =d_2/(D_1-s)
?(= )d_2=d_1 [(D_1-s)/D_1 ]
Similarly, at the next frame capture, if a ring of width d3 is imaged as the outermost projected ring by the camera, then,
d_3=d_2 [(D_2-s)/D_2 ]
=d_1 [(D_1-s)/D_1 ]·[(D_1-2·s)/(D_1-s)]
=d_1 [(D_1-2·s)/D_1 ] and so on.
Hence one can observe that given that D1, D2, D3 etc. are constant for a particular descent, the width of successive rings on the marker, d1, d2, d3 etc., form an arithmetic progression.
Further, it can also be observed that the width of rings is smaller near the center/for the innermost ring, and largest near the marker boundary box/for the outermost ring. Since D1, d1, s are all constants for one flight, this implies that the outer rings, which are wider but constant-spaced, will not overlap unless the rings are relatively thin. Thus, given that the outer periphery of the marker is fixed, this leads to having the marker divided into many rings, whose count is > some threshold. Specifically, this implies that the number of rings is to be = m for no-overlap scenario, then
s=d_1/2·[(D_1-(m-1)·s)/D_1 ]+d_1/2·[(D_1-m·s)/D_1 ]
On simplification, this leads to,
m=(2·D_1·d_1+d_1·s-2·D_1·s)/(2·d_1 )
In an exemplary implementation, a method for determining an exact width ring of the concentric ring marker is provided.
Let the maximum and /or fixed height from which the UAV starts descending vertically be ‘H’. Also let the size of CCD/CMOS backplane of the camera be ‘LxB’ units, where the unit is typically in millimeters. Typically, in normal circumstances, L > B.
Also, let the linear relationship between a ring radius which may be represented by the ring number and the ring width be represented as -
dn = m x n + d0 , where d0 is the width of innermost ring.
Further, as mentioned previously, the rings are constantly spaced. Therefore, let the radii of medial circles of successive rings with R, 2R, 3R… etc.
In the present exemplary implementation, in order to accommodate ‘T’ rings in the concentric ring marker pattern. The value of T is lower bounded as provided below –

T=(2·D_1·d_1+d_1·s-2·D_1·s)/(2·d_1 )
as provided previously, let the value of ‘T’ to be judiciously chosen. Since, the present consideration deals with the projection of outermost ring at all heights. Therefore, the value of ‘T’ must satisfy the following two conditions.
(a) the first condition deals with the projection of Tth ring fitting within the lesser value between ‘L’ and ‘B’ (may be taken as ‘B’).
(b) the second condition deals with a maximally fit ring in projection from height ‘H’, it must be so that projection of any hypothetical (T+1)th ring may not fit within the width ‘B’.
Therefore, the first condition may be expressed as:
H.B/f=T.R+d_T/2
Substituting for dT as m • T + d0, and simplifying, we get –
m=((2.H.B-2.f.R-f.d_0))/((T.f))
Similarly, the second condition is expressed as:
(T+1).R+d_((T+1))/2=(H.B)/fAgain, substituting for dT+1 as m • (T+1) + d0, and simplifying, we get:
m=((2.H.B-f.d_0-2.f.(T+1).R))/((T+1).f)
Thus we have bounded the multiplier m, in the linear relationship dn = m • n + d0, to be between two bounds, as below:
((2.H.B-f.d_0-2.f.(T+1).R))/((T+1).f)=m=((2.H.B-2.f.R-f.d_0))/((T.f))
Therefore, the person skilled in the art may choose a useful value of ‘m’, while being within these bounds.
"The systems and methods are not limited to the specific embodiments described herein. In addition, components of each system and each method can be practiced independently and separately from other components and methods described herein. Each component and method can be used in combination with other components and other methods."
Throughout the description and claims of this complete specification, the word “comprise” and variations of the word, such as “comprising” and “comprises,” means “including but not limited to,” and is not intended to exclude, for example, other additives, components, integers or steps. “Exemplary” means “an example of” and is not intended to convey an indication of a preferred or ideal embodiment. “Such as” is not used in a restrictive sense, but for explanatory purposes.