Omnidirectional planar sensor

Abstract

An omnidirectional planar sensor comprising: a conical reflector 20 having a central cone axis, the conical reflector being configured to reflect radiation, incident on the conical reflector in a direction towards the central cone axis and in a plane perpendicular to the central cone axis, in a direction parallel to the central cone axis within a cylindrical radiation transmission path; a parabolic reflector 40, with an axis of symmetry offset from the central cone axis of the conical reflector, configured to receive radiation reflected from the conical reflector and to concentrate the received radiation in an image plane outside of the cylindrical radiation transmission path; and a detector 60, positioned in the image plane outside of the cylindrical radiation transmission path where the received radiation is concentrated, and configured to detect the received radiation and output image information dependent on the detected received radiation. The camera may me used at infrared or visible frequencies or with a LIDAR. It may be incorporated in an autonomous vehicle or robot.

Description

TECHNICAL FIELD The present disclosure relates to apparatus and methods for object detection and localisation. In particular, but not exclusively, it relates to an omnidirectional planar sensor for detecting objects. Aspects of the invention relate to an omnidirectional planar sensor, a method of detecting an object using an omnidirectional planar sensor, and an autonomous mobile robot having an omnidirectional planar sensor for the detection and tracking of an object. BACKGROUND The detection and tracking of objects may be traditionally carried out using various types of sensors and sensor systems, for example by the use of stereoscopic cameras, RGB-D cameras, radar, or 2D or 3D lidar systems. However, such systems may have various drawbacks in relation to operational effectiveness or cost. For example, stereoscopic cameral and RGB-D cameras have limited range and angle of detection and radar and lidar systems, which can have a larger angle of detection than stereoscopic cameras or RGB-D cameras, can be very expensive and require significant, and therefore costly, processing in order to differentiate objects in an environment, such as the detection of moving objects from a static background in a factory or a street scene for example. Further, in order to provide more robust systems, data from several sensors, such as an RGB-D camera and a 2D lidar system, may be combined, further increasing cost and computational complexity. In order to provide three hundred and sixty degree detection of objects, multiple cameras, lidars or radars, or omnidirectional cameras, lidars or radars may be utilized, though such systems are prohibitively costly in many everyday applications. In order to provide detection of some objects, thermal imaging may be used which can identify objects of different temperature, such that an object may be more easily identified in an environment which is cooler or hotter than the object. Such thermal imaging may be used in the detection of people in an environment. Autonomous mobile robots (AM Rs) have recently gained popularity in industry and are used, for example, to transport components or parts around a factory. In order to do this, such AMRs need to be able to freely navigate locations whilst avoiding obstacles including static and moving objects including people. Therefore, such AMRs may be provided with detection and tracking systems of the types noted above. It is an aim of the present invention to address at least some of the disadvantages associated with the prior art. SUMMARY OF THE INVENTION Aspects and embodiments of the invention provide an omnidirectional planar sensor, a method of detecting an object using an omnidirectional planar sensor, and an autonomous mobile robot having an omnidirectional planar sensor for the detection and tracking of an object via imaging sensors, as claimed in the appended claims. According to an aspect of the invention there is provided an omnidirectional planarsensor comprising: a conic reflector, or conical reflector, having a central cone axis, the conical reflector being configured to reflect radiation, incident on the conical reflector in a direction towards the central cone axis and in a plane perpendicular to the central cone axis, in a direction parallel to the central cone axis within a cylindrical radiation transmission path; a parabolic reflector, with an axis of symmetry offset from the central cone axis of the conical reflector, configured to receive radiation reflected from the conical reflector and to concentrate the received radiation in an image plane outside of the cylindrical radiation transmission path; and a sensor, positioned in the image plane outside of the cylindrical radiation transmission path where the received radiation is concentrated, and configured to detect the received radiation and output image information dependent on the detected received radiation. An advantage of this invention is that detection of objects located in a detection volume can be provided with a simple optical arrangement that is resistant to interference from outside of the detection volume. In particular, but not exclusively, it may use an imaging sensor, such as a thermal sensor or thermal radiation sensor arrangement used to detect thermal radiation. With such a sensor, it is desirable to avoid the detection of any unwanted heat sources, such as ceiling mounted lights and underfloor heating, and the present invention provides for the selection of a detection volume which avoids these areas. The sensor may be configured to detect received radiation of the wavelengths of interest. In particular, but not exclusively, thermal radiation between 8 micrometers and 14 micrometers may be detected by a thermal sensor, visible spectrum radiation between 380 nanometers and 750 nanometers may be detected by cameras or event cameras, near-infrared radiation between 750 nanometers and 2500 nanometers may be detected by lidar or depth camera sensors, and millimetre-wave radiation between 1 millimetre and 10 millimeters may be detected by millimetre-wave radars. The axis of symmetry of the parabolic reflector may be offset angularly from the central cone axis. The axis of symmetry of the parabolic reflector may be offset laterally in a plane perpendicular to the central cone axis. An external reflective surface of the conical reflector and an external reflective surface of the parabolic reflector may comprise materials reflective over the wavelengths of interest. The external reflective surface of the conical reflector and the external reflective surface of the parabolic reflector may comprise a gold outer layer. The surface roughness of the external reflective surface of the conical reflector and the external reflective surface of the parabolic reflector may be less than or equal to 0.8 micrometers. The conical reflector may be spatially separated from the parabolic reflector such that there is no overlap of the conical reflector and the parabolic reflector in a plane perpendicular to the central cone axis. The spatial separation between the conical reflector and the parabolic reflector may be maintained by one or more support column. The one or more support column may be formed of material which is transparent over the wavelengths of interest. According to an aspect of the invention there is provided a method of detecting an object using an omnidirectional planar sensor as set out above, the method comprising: positioning the omnidirectional planar sensor such that the central cone axis of the conical reflector is perpendicular to a detection volume in which objects to be detected are located; receiving, at the conical reflector, radiation from an object in the detection volume, the radiation from the object in the detection volume being incident on the conical reflector in a direction towards the central cone axis; detecting, at the sensor, the received radiation; and outputting image information from the sensor, the image information being dependent on the detected received radiation. The detection volume may extend in a plane extending three hundred and sixty degrees around the conical reflector and may have a height, parallel to the central cone axis, equal to the height of the conical reflector along the central cone axis. The method may comprise: forming the image information from the detected received radiation, the image information providing a three hundred and sixty degree radiation map indicating a location for the object in the detection volume relative to the omnidirectional planar sensor. The method may comprise: normalising the radiation map by identifying a maximum detected radiation value and a minimum detected radiation value from the detected received radiation; and binarising the radiation map using a threshold between the maximum detected radiation value and the minimum detected radiation value to provide pixels relating to the background and pixels relating to the object. The method may comprise: detecting the direction of the object by determining an angle of a centre of mass of pixels relating to the object. The method may comprise: detecting the distance to the object by determining the size of the area of pixels relating to the object. According to an aspect of the invention there is provided an autonomous mobile robot having an omnidirectional planar sensor as set out above, the autonomous mobile robot comprising: a processor to receive image information from the omnidirectional planar sensor and to cause movement of the autonomous mobile robot dependent on the received image information. 4 The autonomous mobile robot may be configured to be guided relative to an object dependent on the detected location of the object relative to the omnidirectional planar sensor. The autonomous mobile robot may be configured to rotate about an axis parallel to the central cone axis of the omnidirectional planar sensor when the object is outside of an angular range centred on a reference axis in the plane perpendicular to the central cone axis. The autonomous mobile robot may be configured to move forwards at a first speed when the object is beyond a first threshold distance from the autonomous mobile robot. The autonomous mobile robot may be configured to move forwards at a second speed, greater than the first speed, when the object is beyond a second threshold distance from the autonomous mobile robot, the second threshold distance from the autonomous mobile robot being greater than the first threshold distance from the autonomous mobile robot. Within the scope of this application, it is expressly intended that the various aspects, embodiments, examples and alternatives set out in the preceding paragraphs, in the claims and / or in the following description and drawings, and in particular the individual features thereof, may be taken independently or in any combination. That is, all embodiments and / or features of any embodiment can be combined in any way and / or combination, unless such features are incompatible. The applicant reserves the right to change any originally filed claim or file any new claim accordingly, including the right to amend any originally filed claim to depend from and / or incorporate any feature of any other claim although not originally claimed in that manner. BRIEF DESCRIPTION OF THE DRAWINGS One or more embodiments of the invention will now be described, by way of example only, with reference to the accompanying drawings, in which: Figure 1 is a cross section of an omnidirectional planar sensor, according to an embodiment of the invention, illustrating ray tracing of radiation interacting with the components of the omnidirectional planar sensor; Figure 2 illustrates a perspective view of an omnidirectional planar sensor and a resultant output image according to an embodiment of the invention; Figure 3 is a flow diagram of a method of detecting an object using an omnidirectional planar sensor according to an embodiment of the invention; Figure 4 illustrates a side view of an autonomous mobile robot having an omnidirectional planar sensor according to an embodiment of the invention; and Figure 5 illustrates segmentation of the area around an autonomous mobile robot for instructing movement of the autonomous mobile robot. In the drawings, like parts are denoted by like reference numerals. DETAILED DESCRIPTION Examples of the present disclosure relate to an omnidirectional planar sensor. In particular, examples of the present disclosure relate to an omnidirectional planar sensor for detecting objects via imaging. Various herein described examples of the present disclosure relate to an omnidirectional planar thermal sensor for detecting objects, such as a human body, via thermal imaging, though the person skilled in the art will understand that radiation of different wavelengths, other than those defining a thermal or infrared spectrum, may be detected by the omnidirectional planar sensor, in order to detect objects. For example, the sensor may be configured to detect thermal radiation between 8 micrometers and 14 micrometers using a thermal sensor, visible spectrum radiation between 380 nanometers and 750 nanometers using cameras or event cameras, near-infrared radiation between 750 nanometers and 2500 nanometers using lidar or depth camera sensors, and millimetre-wave radiation between 1 millimetre and 10 millimeters using millimetre-wave radars. Non-limiting examples will now be described with reference to the accompanying drawings. The figures illustrate an omnidirectional planar sensor 10, a method 300 of detecting an object using an omnidirectional planar sensor 10, and an autonomous mobile robot 100 having an omnidirectional planar sensor 10. Figure 1 illustrates an omnidirectional planar sensor 10 comprising a conical reflector 20, a parabolic reflector 40, and a sensor 60. In thermal sensing applications the omnidirectional planar sensor 10 may be an omnidirectional planar thermal sensor 10 and the sensor 60 may be a thermal sensor 60. The conical reflector 20 has a central cone axis 22 extending from the centre of the base 24 of the conical reflector 20 to the vertex 26 or apex of the conical reflector 20. The conical reflector 20 is configured or arranged to reflect radiation, such as thermal radiation, which is incident on the conical outer surface 30 of the conical reflector 20 in a direction towards the central cone axis 22 and in a plane perpendicular to the central cone axis 22, in a direction which is parallel to the central cone axis 22 within a cylindrical radiation transmission path 28. A number of ray traces 32 in planes perpendicular to the central cone axis 22 are shown in Figure 1 to illustrate the path of the incoming radiation and its reflection throughout the omnidirectional planar thermal sensor 10. In the example of a thermal sensing arrangement, thermal radiation may include infrared radiation between wavelengths of 1 mm to 700 nanometres. In particular, in an omnidirectional planar thermal sensor 10 which is intended to detect people in an environment, that is to detect a human body with a temperature in the region of 36 °C to 38 °C (96.8 °F to 100.4 °F or 309.15 Kelvin to 311.15 Kelvin), the wavelength of the thermal radiation of interest is limited to long wavelength infrared between 8 and 14 micrometres. The parabolic reflector 40, has an axis of symmetry 42 which is offset from the central cone axis 22 of the conical reflector 20. In the example shown in Figure 1, the axis of symmetry 42 of the parabolic reflector 40 is offset angularly from the central cone axis 22. That is, the axis of symmetry 42 of the parabolic reflector 40 is offset at an angle from the central cone axis 22 of the conical reflector 20. It will be understood that in other embodiments, the axis of symmetry 42 of the parabolic reflector 40 may alternatively, or additionally, be offset laterally in a plane perpendicular to the central cone axis 22, with the purpose that the radiation reflected from the parabolic reflector 40 is directed to be focused on an image plane 62 outside of the cylindrical radiation transmission path 28. The parabolic reflector 40 is configured or arranged to receive radiation which has been received at the conical outer surface 30 of the conical reflector 20 in a direction towards the central cone axis 22 and in a plane perpendicular to the central cone axis 22 and subsequently reflected from the conical outer surface 30 of the conical reflector 20. The parabolic reflector 40 is configured or arranged to receive the radiation at a parabolic reflector outer surface 50 of the parabolic reflector 40, and to concentrate the received radiation in the image plane 62 outside of the cylindrical radiation transmission path 28. The sensor 60 is positioned or located in the image plane 62 outside of the cylindrical radiation transmission path 28 where the received radiation is concentrated. Therefore, the sensor 60 does not interfere with the path of the radiation which is reflected from the conical reflector 20 to the parabolic reflector 40. The sensor 60 is configured or arranged to detect the received radiation and output image information dependent on the detected received radiation. The sensor 60 may be a thermal sensor 60 configured to detect received thermal radiation of wavelengths between 8 micrometers and 14 micrometers, thereby limiting detection to long-wavelength infrared, often termed thermal infrared, where passive detection of an object which is only slightly different in temperature from the background temperature is possible. In such a thermal sensor 60, a person may be detectable relative to a background when the temperature of the person is only slightly different from the temperature of the background. Thermal sensors 60 which may be used to detect thermal radiation of wavelengths between 8 micrometers and 14 micrometers include thermopile sensors or microbolometers. The external reflective surface of the conical reflector 20, that is the conical outer surface 30 of the conical reflector 20, and the external reflective surface of the parabolic reflector 40, that is the parabolic reflector outer surface 50 of the parabolic reflector 40, may comprise material or materials reflective over the wavelengths of interest, in order to limit losses in the transmission of the radiation to the image plane 62. For example, the external reflective surface of the conical reflector 20, that is the conical outer surface 30 of the conical reflector 20, and the external reflective surface of the parabolic reflector 40, that is the parabolic reflector outer surface 50 of the parabolic reflector 40, may comprise infrared reflective material in order to limit losses in the transmission of the thermal radiation to the image plane 62, when the apparatus is intended to detect thermal radiation. The conical outer surface 30 of the conical reflector 20 and the parabolic reflector outer surface 50 of the parabolic reflector 40 may be formed of, or coated in, for example, gold, silver, aluminium, copper, nickel, steel, or other infrared reflective material. In some embodiments the conical reflector 20 and / or the parabolic reflector 40 may be made of steel or other infrared reflective material. In one example, the reflector components of the system may be formed using a steel alloy such as grade 410 stainless steel, which is a martensitic alloy containing a high percentage of chromium. This material is a good reflector of infrared radiation, particularly in the wavelength band between 8 and 14 micrometres where passive detection of an object which is only slightly different in temperature from the background temperature is possible. Grade 410 stainless steel has high corrosion resistance, ensuring long-lasting reflection even if the reflective surfaces are not protected, and is a relatively hard alloy compared to other metals, thereby affording the reflective surfaces good resistance against scratches and general handling. In other embodiments, the conical reflector 20 and / or the parabolic reflector 40 may be formed by coating the surfaces, which are intended to be reflective surfaces, of a substrate, such as a polymer based substrate, with a thin layer of reflective material. In particular, the conical reflector 20 and / or the parabolic reflector 40 may be formed by coating the surfaces, which are intended to be reflective surfaces, of a substrate, such as a polymer based substrate, with a thin layer of infrared reflective material such as gold, silver, aluminium, copper, or nickel. For example, in some embodiments the conical reflector 20 is formed of a conical shaped substrate, such as a conical based polymer substrate, and the parabolic reflector 40 is formed of a parabolic substrate, such as a parabolic shaped polymer substrate. The conical shaped substrate and the parabolic shaped substrate may then be coated with, for example, a gold outer layer to provide infrared reflecting outer surfaces. A gold coating provides both a good infrared reflective surface and resistance to oxidation, and therefore may provide a long term consistent infrared reflective surface, when compared to some other coating materials which may oxidise and therefore cause instrumental drift over time. In order to ensure specular reflection of the radiation, the external reflective surface of the conical reflector 20, that is the conical outer surface 30 of the conical reflector 20, and the external reflective surface of the parabolic reflector 40, that is the parabolic reflector outer surface 50 of the parabolic reflector 40, should be at least an order of magnitude smoother than the wavelength of the radiation to be reflected. When the wavelength of interest is between 8 micrometers and 14 micrometers, that is the long wavelength infrared used for the passive imaging of objects which may be only slightly different in temperature than the surrounding environment, for example the human body, the surface roughness of the reflective components of the omnidirectional planar thermal sensor 10, at least the external reflective surface of the conical reflector 20 and the external reflective surface of the parabolic reflector 40, should be at least an order of magnitude smoother than the wavelength of the thermal radiation to be reflected. In particular, the external reflective surface of the conical reflector 20 and the external reflective surface of the parabolic reflector 40, should be less than or equal to 0.8 micrometers. In some embodiments the surface roughness of those reflective surfaces may be 0.4 micrometers for example. As can be seen in Figure 1, the conical reflector 20 is spatially separated from the parabolic reflector 40 such that there is no overlap of the conical reflector 20 and the parabolic reflector 40 in a plane perpendicular to the central cone axis 22. In this way when the radiation is reflected from the parabolic reflector 40 towards the image plane 62, the conical reflector 20 does not impinge in the transmission path of the radiation. The spatial separation between the conical reflector 20 and the parabolic reflector 40 may be maintained by one or more support column 36. Ideally the one or more support column 36 is formed of a material that is transparent over the wavelengths of interest, such that the support columns 36 do not interfere with the path of radiation from an object to be detected towards the conical reflector 20. In embodiments where thermal radiation is being detected, ideally the one or more support column 36 is formed of an infrared transparent material, such that the support columns 36 do not interfere with the path of thermal radiation from an object to be detected towards the conical reflector 20. In some embodiments the one or more support column 36 is formed of a material which is transparent, or substantially transparent, in the range of wavelengths between 8 micrometers and 14 micrometers. In alternative embodiments, the one or more support column 36 may be formed of a material which is not transparent or substantially transparent to radiation, in particular a material which is not transparent or substantially transparent to thermal radiation between the wavelengths of 8 micrometers and 14 micrometers when the omnidirectional planar thermal sensor 10 is intended to detect people in an environment. In such arrangements, when the one or more support column 36 is not transparent or substantially transparent in the wavelengths of interest, any reflections caused by the one or more support column 36 may be removed computationally from image information which is output by the sensor 60. Other parts of the omnidirectional planar sensor 10 that may cause reflections may also be formed of material which is transparent or substantially transparent in the range of wavelengths of interest, for example between 8 micrometers and 14 micrometers, or reflections from such other parts of the omnidirectional planar sensor 10 can be removed computationally. Figure 2 illustrates an omnidirectional planar thermal sensor 10 relative to an object 80, in this case a person, which is located in a detection volume 70, the detection volume 70 being defined by the geometry of the conical reflector 20. Figure 2 also illustrates image information 90 which is formed from the detected received thermal radiation, the image information 90 providing a three hundred and sixty degree thermal map indicating a location for the object 80 in the detection volume 70 relative to the omnidirectional planar thermal sensor 10. It will be noted that in other examples, the omnidirectional planar sensor 10 may be detecting other wavelengths of interest, potentially to detect objects other than a person. Image information 90 may be formed from the detected received radiation to provide a three hundred and sixty degree radiation map indicating a location for the object 80 in the detection volume 70 relative to the omnidirectional planar sensor 10. In the example of the omnidirectional planar thermal sensor 10, the image information 90 provides the thermal map in the form of a two-dimensional function where f is the temperature value, and u and v are spatial coordinates in the thermal map. To calculate the projection of a given real-world point to the image plane of the thermal map, the image plane centre is aligned with the geometric centre of the parabolic reflector 40. With the arrangement presented in Figure 2, the detectable three-dimensional points, WP= [X,y,Z]r, are constrained between the horizontal planes of the upper and lower height of the conical reflector 20. The projection of the real-world point to the image plane of the thermal map is described in Equation 1 below. f(u,vy = [r-cos(6),r-sin(6)] Equation 1 Where parameters r and 0 can be calculated using Equation 2 and Equation 3 below. 0 = atan2(Y,X) Equation 2 r = V(ru2+rv2) Equation 3 Then, parameter r can be decomposed in ru and rv. These values depend on the height of the three-dimensional point as described in Equation 4 and Equation 5 below. „ I Z—Z^ \ usensor ,- .. , r,, = (------I-- Equation 4 u Vz^-zJ 2 Equations \Z2 Z-^ / 2 Where the parameters Zi and Z2 are the lower and upper height limits of the conical reflector. The values sensor and vsenSor are the horizontal and vertical resolution of the thermal sensor respectively. Due to the sensor being placed in an offset position, Z2 is different for the upper or lower half of the image. However, because there is a limited height difference, it can be approximated without significant impact on the thermal map projection. As the equations show, because of the optical arrangement, points can only be detected on a set of planes, which, in some embodiments, are parallel to the floor. Because of this, instead of using shape features, as may be required when using cameras with image recognition algorithms, temperature measurements represented by pixel intensity values / (u,v)T, are employed to detect objects 80, such as people. In the scenario where the omnidirectional 12 planar thermal sensor 10 is configured to detect people, it is noted that some external factors, such as clothing or environmental temperature may affect the apparent temperature of a person. For example, due to such environmental factors, detected temperatures for a person can range from 22°C to 35°C, and so such environmental factors may be taken into consideration when determining any threshold temperature used for detecting people in the environment. Figure 3 illustrates a flow diagram of a method 300 of detecting an object 80 using an omnidirectional planar sensor 10. In particular, the method 300 of detecting an object 80 using an omnidirectional planar sensor 10 can be carried out using the omnidirectional planar sensor 10, such as the omnidirectional planar thermal sensor 10, as previously described. Some blocks of the method 300 may be carried out using processing means, such as a processor or processing circuitry integral with, or operably connected to, the omnidirectional planar sensor 10. The method 300 comprises block 302 where the omnidirectional planar sensor 10 is positioned such that the central cone axis 22 of the conical reflector 20 is perpendicular to a detection volume 70 in which objects 80 to be detected are, at least in part, located. The detection volume 70 extends in, or parallel to, a plane extending three hundred and sixty degrees around the conical reflector 20. The plane is perpendicular to the central cone axis 22. The detection volume 70 has a height, parallel to the central cone axis 22, equal to, or substantially equal to, the height of the conical reflector 20 along the central cone axis 22. At block 304 the method 300 comprises: receiving, at the conical reflector 20, radiation from an object 80 in the detection volume 70, the radiation from the object 80 in the detection volume 70 being incident on the conical reflector 20 in a direction towards the central cone axis 22. When detecting thermal radiation, the received thermal radiation is indicative of the temperature of the object 80 from which the thermal radiation emanates. At block 306 the received radiation is detected at the sensor 60. The sensor 60 is positioned in the image plane 62 outside of the path of the radiation between the conical reflector 20 and the parabolic reflector 40. Radiation is therefore received from three hundred and sixty degrees around the omnidirectional planar sensor 10, with different radiation being received dependent on the 13 object 80 from which the radiation emanates at any given angle around the omnidirectional planar sensor 10. For example, different thermal radiation can be received dependent on the object 80 from which the thermal radiation emanates at any given angle around an omnidirectional planar thermal sensor 10. Therefore, multiple objects of interest may be detected at the same time. The received radiation has an intensity value dependent on the radiation detected, and an angular value, the angular value being dependent on the angle of incidence, in the detection volume 70, at which the received radiation is incident on the conical reflector 20 in a direction towards the central cone axis 22. In the example of the omnidirectional planar thermal sensor 10 received thermal radiation has an intensity value dependent on the temperature detected, and an angular value, the angular value being dependent on the angle of incidence, in the detection volume 70, at which the received thermal radiation is incident on the conical reflector 20 in a direction towards the central cone axis 22. At block 308 image information 90 is formed from the detected received radiation, the image information 90 providing a three hundred and sixty degree radiation map indicating a location for the object 80 in the detection volume 70 relative to the omnidirectional planar sensor 10. The radiation map is formed of pixels, each pixel having an intensity which is indicative of the radiation from an object 80 from which the radiation has been received, the radiation map further providing information relating to the location of the object 80 in the detection volume 70. In the example of the omnidirectional planar thermal sensor 10, image information 90 is formed from the detected received thermal radiation, the image information 90 providing a three hundred and sixty degree thermal map indicating a location for the object 80 in the detection volume 70 relative to the omnidirectional planar thermal sensor 10. The thermal map is formed of pixels, each pixel having an intensity which is indicative of the temperature of an object 80 from which the thermal radiation has been received, the thermal map further providing information relating to the location of the object 80 in the detection volume 70. At block 310 image information 90 is output from the sensor 60, the image information 90 being dependent on the detected received radiation. In the example of the omnidirectional planar thermal sensor 10, when detecting a particular object 80, such as a person, in an environment, that object may be at a temperature which contrasts with environmental temperature. For example, the body temperature of a person generally contrasts with the surrounding environmental temperature. In other examples, the radiation value from an object may be seen to differ or contrast with radiation values from the environment. The image information may be segmented by an algorithm that classifies pixels as an object of interest, such as a person, or background. To do so, the radiation map may be firstly normalised, to enhance the detection range. Therefore, optionally, in block 312 the radiation map may be normalised by identifying a maximum detected radiation value and a minimum detected radiation value from the detected received radiation. The maximum detected radiation value and minimum detected radiation value can then be used to delimit the boundary of the radiation value range for detection. Then, a circular mask may be applied to provide equal readings in all directions and to prevent interference from the edges of the radiation map. In the example of the omnidirectional planar thermal sensor 10, the thermal map may be firstly normalised, to enhance the detection range. Therefore, optionally, in block 312 the thermal map may be normalised by identifying a maximum detected temperature and a minimum detected temperature from the detected received thermal radiation. The maximum detected temperature and minimum detected temperature can then be used to delimit the boundary of the temperature range for detection. Then, a circular mask may be applied to provide equal readings in all directions and to prevent interference from the edges of the thermal map. The minimum temperature is identified as the lowest value in the thermal map and the highest temperature is identified as the highest value in the thermal map. In a greyscale image this would equate to the minimum detected temperature being set as a black pixel and the maximum detected temperature being set as a white pixel, or given respective values of 0 and 255 when a range of values from 0 to 255 is provided. All other temperatures in between the minimum and maximum values are greyscale values therebetween, that is values between 0 and 255. In some embodiments, in particular when detecting a person in an environment, a different arrangement can be provided where a greater resolution in the thermal map may be provided closer to an expected temperature of the object, for example 36 °C to 38 °C when the object is a person. In such an arrangement, any temperatures lower than 26 °C may be 15 set as black, or having a value of 0, and any temperature higher than 38 °C may be set as white, or having a value of 255, with all other temperatures in between being a greyscale value, that is a value between 0 and 255, therebetween. Optionally, in block 314 the radiation map may be binarised using a threshold between the maximum detected radiation value and the minimum detected radiation value to provide pixels relating to the background and pixels relating to the object 80. For example, the thermal map for the omnidirectional planar thermal sensor 10 may be binarised using a threshold between the maximum detected temperature and the minimum detected temperature to provide pixels relating to the background and pixels relating to the object 80. An example of a technique for binarisation is the use of Otsu’s method, an algorithm which can be used to provide a global adaptive threshold, tgi, separating the pixels into two classes, foreground and background, using a probabilistic method. If the object temperature is expected to be greater than that of the ambient environmental temperature, the binarisation should be applied as shown in Equation 6 below. (1 f (’ll, v)T >tai b(u, v) = { T 5 Equation 6 [o f(u,vy <tgl The result is a binarised image where people are highlighted as white circular sectors on a black background. However, other known segmentation methods may be applied to identify the intensity threshold. By the process of binarisation, the background can be presented as black pixels, that is having a value 0, and an object 80, such as a person, can be presented as white pixels, that is having a value 1. In particular, such normalisation and binarisation may be ideal for identifying people in an environment where the environmental temperature and other objects in that environment are lower than the temperature of the person. Optionally, at block 316, the direction of the object 80 may be detected by determining an angle of a centre of mass of pixels relating to the object 80. That is, the centre of mass of the white pixels, or pixels having a value of 1, when the radiation or thermal map is binarised, can be used to determine an angle at which the object 80 is positioned in the detection volume 70. 16 Optionally, at block 318, the distance to the object 80 may be detected by determining the size of the area of pixels relating to the object 80. That is, the area of the white pixels, or pixels having a value of 1, when the radiation or thermal map is binarised, can be used to determine the distance to the object 80 in the detection volume 70. The area on the radiation or thermal map will increase in size, for a given object 80, when that object 80 moves closer to the omnidirectional planar sensor 10 and decreases as the object 80 moves further away from the omnidirectional planar sensor 10. To obtain distance estimates, using such an omnidirectional planar sensor 10, regression analysis is used to correlate the size of the area of pixels, or regions, on the radiation or thermal map with the separation between the omnidirectional planar sensor 10 and the object 80. An example of such a model is a two-term exponential model, described below by Equation 7. r(p)=a ■ eb'p + c • ed'p Equation 7 Where r is the estimated distance to the object, p is the size of the area of pixels, and a, b, c, and d are coefficients that are empirically obtained. Finally, in static applications, when objects 80, such as people, are moving, the detected area may be reduced because the omnidirectional planar sensor 10 detects the lateral profile of the object 80. Because of this, a weighting coefficient can be applied to indicate that the object 80 is closer than the area indicates. However, depth estimations based on the size of the area of pixels may be unreliable due to various reasons. For example, when detecting people, a person who is not facing the omnidirectional planar sensor 10 will present differently to the same person when facing the omnidirectional planar sensor 10, and therefore result in a different size of the area of pixels on the radiation or thermal map. Further, different people can have different sizes which may also impact the distance estimates. Additionally, partial occlusions, such as partial thermal occlusions could impact the distance estimates. Whilst detection of an objects angular position and distance can be provided using the method above, in some embodiments, in order to improve the location detection of an object 80, the system may provide fusion of data from multiple sensors, such as the fusion of data from an 17 omnidirectional planar thermal sensor 10 and data from a lidar, such as a two-dimensional lidar, in a multi-modal detection system. In this way the distance determination for an object 80 may be improved. In some embodiments, the thermal sensing from the omnidirectional planar thermal sensor 10 may be used to determine the direction of the object 80 in the detection volume 70 and the lidar may be used to determine the distance to the object 80 in the detection volume 70. In one arrangement, objects 80 are detected first using the omnidirectional planar thermal sensor 10 and then the distance to the object 80 is determined using the lidar. In another arrangement, objects 80 are detected, and their distance determined, first using a lidar and then false positive detections, such as the detection of an object 80 which is not a person, are filtered out using the omnidirectional planar thermal sensor 10. In another arrangement, objects 80 are independently detected and the distance to them determined using the omnidirectional planar thermal sensor 10 and a lidar, with both sets of data then being fused using a sensor fusion mechanism, such as the Fresner-Potter smoother. Figure 4 illustrates one example of a use case for the omnidirectional planar sensor 10. In this case, an autonomous mobile robot 100 is provided with an omnidirectional planar sensor 10, and optionally other sensors such as a lidar. The autonomous mobile robot 100 comprises processing means, in the form of a processor 102 or processing circuitry to receive image information from the omnidirectional planar sensor 10 and to cause movement of the autonomous mobile robot 100 dependent on the received image information. The autonomous mobile robot 100 may be configured to be guided relative to an object 80, such as a person, dependent on the detected location of the object 80 relative to the omnidirectional planar sensor 10. Figure 5 illustrates an example of the use of object location parameter limits to provide guidance for an autonomous mobile robot 100. In the arrangement of Figure 5, the surroundings of the autonomous mobile robot 100 are segmented into areas in which, when the object 80 is located therein, the autonomous mobile robot 100 is configured to carry out particular movement instructions. For example, the autonomous mobile robot 100 is configured to rotate about an axis parallel to the central cone axis 22 of the omnidirectional planar sensor 10 when the object 80 is outside of an angular range 110 centred on a reference axis 120 in the plane perpendicular to the central cone axis 22. Therefore, if the detected object 80 of interest is in the regions 112, 114 outside of the 18 angular range 110 centred on a reference axis 120, then the autonomous mobile robot 100 is configured to rotate, in particular such that the reference axis 120 is directed towards the object 80 or moved towards the object 80 in such a way to be more angularly correlated to the object 80. The autonomous mobile robot 100 may be configured to move forwards at a first speed when the object 80 is beyond a first threshold distance 116 from the autonomous mobile robot 100. Alternatively, the autonomous mobile robot 100 may be configured to move forwards at a speed which is proportional to the distance to the object 80, when the object 80 is beyond a first threshold distance 116 from the autonomous mobile robot 100. The autonomous mobile robot 100 may be configured to move forwards at a first speed when the object 80 is beyond a first threshold distance 116 from the autonomous mobile robot 100 but less than a second threshold distance 118 from the autonomous mobile robot 100, or to move at a speed which is proportional to the distance to the object 80, when the object 80 is beyond a first threshold distance 116 from the autonomous mobile robot 100 but less than a second threshold distance 118 from the autonomous mobile robot 100. The autonomous mobile robot 100 may be configured to move forwards at a second speed, greater than the first speed, when the object 80 is beyond the second threshold distance 118 from the autonomous mobile robot 100, the second threshold distance 118 from the autonomous mobile robot 100 being greater than the first threshold distance 116 from the autonomous mobile robot 100. The second speed may be a maximum speed of the autonomous mobile robot 100. In an example, an autonomous mobile robot 100 may be configured to be stationary, that is have a zero velocity, when an object 80, such as a person, is closer than 1.5 metres to the autonomous mobile robot 100. If the object 80 is within the range of 1.5 metres to 2.5 metres, the autonomous mobile robot 100 will move forward at a speed proportional to the distance, that is, move at a higher speed when at a further distance from the object 80. Lastly, if the distance to the object 80 is greater than 2.5 metres, the autonomous mobile robot 100 will advance at a maximum speed of 0.35 m / s. Although embodiments of the present invention have been described in the preceding paragraphs with reference to various examples, it should be appreciated that modifications to the examples given can be made without departing from the scope of the invention as claimed. 19 The omnidirectional planar sensor 10 and methods 300 described herein can be adapted for use with different applications and to detect different objects. The omnidirectional planar sensor 10 and methods 300 described herein can be used, for example, in applications such as autonomous mobile robot control, safety systems where for example heavy equipment may be moving alongside personnel, and security systems where the detection of personnel may be required in order to trigger the operation of alarms or visual camera systems. Of course, the omnidirectional planar sensor 10 and methods 300 described herein are not to be limited to such applications. Features described in the preceding description may be used in combinations other than the combinations explicitly described. Although functions have been described with reference to certain features, those functions may be performable by other features whether described or not. Although features have been described with reference to certain embodiments, those features may also be present in other embodiments whether described or not. Whilst endeavoring in the foregoing specification to draw attention to those features of the invention believed to be of particular importance it should be understood that the Applicant claims protection in respect of any patentable feature or combination of features hereinbefore referred to and / or shown in the drawings whether or not particular emphasis has been placed thereon.

Claims

1. An omnidirectional planar sensor comprising:a conical reflector having a central cone axis, the conical reflector being configured to reflect radiation, incident on the conical reflector in a direction towards the central cone axis and in a plane perpendicular to the central cone axis, in a direction parallel to the central cone axis within a cylindrical radiation transmission path;a parabolic reflector, with an axis of symmetry offset from the central cone axis of the conical reflector, configured to receive radiation reflected from the conical reflector and to concentrate the received radiation in an image plane outside of the cylindrical radiation transmission path; anda sensor, positioned in the image plane outside of the cylindrical radiation transmission path where the received radiation is concentrated, and configured to detect the received radiation and output image information dependent on the detected received radiation.

2. An omnidirectional planar thermal sensor according to claim 1, wherein the sensor is a thermal sensor configured to detect received radiation of wavelengths between 8 micrometers and 14 micrometers.

3. An omnidirectional planar sensor according to claim 1, wherein the sensor is a camera configured to detect received radiation of wavelengths between 380 nanometers and 750 nanometers.

4. An omnidirectional planar sensor according to claim 1, wherein the sensor is a lidar configured to detect received radiation of wavelengths between 750 nanometers and 2500 nanometers.

5. An omnidirectional planar sensor according to claim 1, wherein the sensor is a millimetre-wave radar configured to detect received radiation of wavelengths between 1 millimetre and 10 millimetres.

6. An omnidirectional planar sensor according to any preceding claim, wherein the axis of symmetry of the parabolic reflector is offset angularly from the central cone axis.

7. An omnidirectional planar sensor according to any preceding claim, wherein the axis of symmetry of the parabolic reflector is offset laterally in a plane perpendicular to the central cone axis.

8. An omnidirectional planar sensor according to any preceding claim, wherein an external reflective surface of the conical reflector and an external reflective surface of the parabolic reflector comprise materials reflective over the wavelengths of interest.

9. An omnidirectional planar sensor according to claim 8 wherein the external reflective surface of the conical reflector and the external reflective surface of the parabolic reflector comprise a gold outer layer.

10. An omnidirectional planar sensor according to any claim 8 or claim 9, wherein the surface roughness of the external reflective surface of the conical reflector and the external reflective surface of the parabolic reflector is less than or equal to 0.8 micrometers.

11. An omnidirectional planar sensor according to any preceding claim, wherein the conical reflector is spatially separated from the parabolic reflector such that there is no overlap of the conical reflector and the parabolic reflector in a plane perpendicular to the central cone axis.

12. An omnidirectional planar sensor according to claim 11, wherein the spatial separation between the conical reflector and the parabolic reflector is maintained by one or more support column, the one or more support column being formed of a material that is transparent over the wavelengths of interest.

13. A method of detecting an object using an omnidirectional planar sensor as claimed in any of claims 1 to 12, the method comprising:positioning the omnidirectional planar sensor such that the central cone axis of the conical reflector is perpendicular to a detection volume in which objects to be detected are located;receiving, at the conical reflector, radiation from an object in the detection volume, the radiation from the object in the detection volume being incident on the conical reflector in a direction towards the central cone axis;detecting, at the sensor, the received radiation; and22outputting image information from the sensor, the image information being dependent on the detected received radiation.

14. A method according to claim 13, wherein the detection volume extends in a plane extending three hundred and sixty degrees around the conical reflector and has a height, parallel to the central cone axis, equal to the height of the conical reflector along the central cone axis.

15. A method according to claim 13 or claim 14, comprising:forming the image information from the detected received radiation, the image information providing a three hundred and sixty degree radiation map indicating a location for the object in the detection volume relative to the omnidirectional planar sensor.

16. A method according to claim 15, comprising:normalising the radiation map by identifying a maximum detected radiation value and a minimum detected radiation value from the detected received radiation; andbinarising the radiation map using a threshold between the maximum detected radiation value and the minimum detected radiation value to provide pixels relating to the background and pixels relating to the object.

17. A method according to claim 16, comprising:detecting the direction of the object by determining an angle of a centre of mass of pixels relating to the object.

18. A method according to claim 16 or claim 17, comprising:detecting the distance to the object by determining the size of the area of pixels relating to the object.

19. An autonomous mobile robot having an omnidirectional planar sensor as claimed in any of claims 1 to 12, comprising:a processor to receive image information from the omnidirectional planar sensor and to cause movement of the autonomous mobile robot dependent on the received image information.

20. An autonomous mobile robot as claimed in claim 19, wherein the autonomous mobile robot is configured to be guided relative to an object dependent on the detected location of the object relative to the omnidirectional planar sensor.

21. An autonomous mobile robot as claimed in claim 20, wherein the autonomous mobile robot is configured to rotate about an axis parallel to the central cone axis of the omnidirectional planar sensor when the object is outside of an angular range centred on a reference axis in the plane perpendicular to the central cone axis.

22. An autonomous mobile robot as claimed in claim 20 or claim 21, wherein the autonomous mobile robot is configured to move forwards at a first speed when the object is beyond a first threshold distance from the autonomous mobile robot.

23. An autonomous mobile robot as claimed in claim 22, wherein the autonomous mobile robot is configured to move forwards at a second speed, greater than the first speed, when the object is beyond a second threshold distance from the autonomous mobile robot, the second threshold distance from the autonomous mobile robot being greater than the first threshold distance from the autonomous mobile robot.AMENDMENTS TO THE CLAIMS HAVE BEEN FILED AS FOLLOWSCLAIMS1. An omnidirectional planar sensing device comprising:a conical reflector having a central cone axis, the conical reflector being configured to reflect radiation, incident on the conical reflector in a direction towards the central cone axis and in a plane perpendicular to the central cone axis, in a direction parallel to the central cone axis within a cylindrical radiation transmission path;a parabolic reflector, with an axis of symmetry offset from the central cone axis of the conical reflector, configured to receive radiation reflected from the conical reflector and to concentrate the received radiation in an image plane outside of the cylindrical radiation transmission path; anda sensor, positioned in the image plane outside of the cylindrical radiation transmission path where the received radiation is concentrated, and configured to detect the received radiation and output image information dependent on the detected received radiation.

2. An omnidirectional planar sensing device according to claim 1, wherein the sensor is a thermal sensor configured to detect received radiation of wavelengths between 8 micrometers and 14 micrometers.

3. An omnidirectional planar sensing device according to claim 1, wherein the sensor is a camera configured to detect received radiation of wavelengths between 380 nanometers and 750 nanometers.

4. An omnidirectional planar sensing device according to claim 1, wherein the sensor is a lidar configured to detect received radiation of wavelengths between 750 nanometers and 2500 nanometers.

5. An omnidirectional planar sensing device according to claim 1, wherein the sensor is a millimetre-wave radar configured to detect received radiation of wavelengths between 1 millimetre and 10 millimetres.

6. An omnidirectional planar sensing device according to any preceding claim, wherein the axis of symmetry of the parabolic reflector is offset angularly from the central cone axis.

7. An omnidirectional planar sensing device according to any preceding claim, wherein the axis of symmetry of the parabolic reflector is offset laterally in a plane perpendicular to the central cone axis.

8. An omnidirectional planar sensing device according to any preceding claim, wherein an external reflective surface of the conical reflector and an external reflective surface of the parabolic reflector comprise materials reflective over the wavelengths of interest.

9. An omnidirectional planar sensing device according to claim 8 wherein the external reflective surface of the conical reflector and the external reflective surface of the parabolic reflector comprise a gold outer layer.

10. An omnidirectional planar sensor according to any preceding claim, wherein the conical reflector is spatially separated from the parabolic reflector such that there is no overlap of the conical reflector and the parabolic reflector in a plane perpendicular to the central cone axis.

11. An omnidirectional planar sensor according to claim 10, wherein the spatial separation between the conical reflector and the parabolic reflector is maintained by one or more support column, the one or more support column being formed of a material that is transparent over the wavelengths of interest.

12. A method of detecting an object using an omnidirectional planar sensing device as claimed in any of claims 1 to 11, the method comprising:positioning the omnidirectional planar sensing device such that the central cone axis of the conical reflector is perpendicular to a detection volume in which objects to be detected are located;receiving, at the conical reflector, radiation from an object in the detection volume, the radiation from the object in the detection volume being incident on the conical reflector in a direction towards the central cone axis;detecting, at the sensor, the received radiation; andoutputting image information from the sensor, the image information being dependent on the detected received radiation.

13. A method according to claim 12, wherein the detection volume extends in a plane extending three hundred and sixty degrees around the conical reflector and has a height, parallel to the central cone axis, equal to the height of the conical reflector along the central cone axis.

14. A method according to claim 12 or claim 13, comprising:forming the image information from the detected received radiation, the image information providing a three hundred and sixty degree radiation map indicating a location for the object in the detection volume relative to the omnidirectional planar sensing device.

15. A method according to claim 14, comprising:normalising the radiation map by identifying a maximum detected radiation value and a minimum detected radiation value from the detected received radiation; andbinarising the radiation map using a threshold between the maximum detected radiation value and the minimum detected radiation value to provide pixels relating to the background and pixels relating to the object.

16. A method according to claim 15, comprising: detecting the direction of the object by determining an angle of a centre of mass of pixels relating to the object.

17. A method according to claim 15 or claim 16, comprising: detecting the distance to the object by determining the size of the area of pixels relating to the object.

18. An autonomous mobile robot having an omnidirectional planar sensing device as claimed in any of claims 1 to 11, comprising:a processor to receive image information from the omnidirectional planar sensing device and to cause movement of the autonomous mobile robot dependent on the received image information.

19. An autonomous mobile robot as claimed in claim 18, wherein the autonomous mobile robot is configured to be guided relative to an object dependent on the detected location of the object relative to the omnidirectional planar sensing device.

20. An autonomous mobile robot as claimed in claim 19, wherein the autonomous mobile robot is configured to rotate about an axis parallel to the central cone axis of the omnidirectional planar sensing device when the object is outside of an angular range centred on a reference axis in the plane perpendicular to the central cone axis.

521. An autonomous mobile robot as claimed in claim 19 or claim 20, wherein the autonomous mobile robot is configured to move forwards at a first speed when the object is beyond a first threshold distance from the autonomous mobile robot.10 22. An autonomous mobile robot as claimed in claim 20, wherein the autonomous mobilerobot is configured to move forwards at a second speed, greater than the first speed, when the object is beyond a second threshold distance from the autonomous mobile robot, the second threshold distance from the autonomous mobile robot being greater than the first threshold distance from the autonomous mobile robot.15

Images