Method and apparatus for sensing a visible light target

Abstract

The application provides a visible light target sensing method and device, applied to the technical field of industrial Internet of Things, determining the receiving signal strength of each receiving end, wherein the transmitted signal is visible light, realizing integrated sensing of the target through directional radiation and light intensity consistency of the visible light, realizing mutual assistance of the positioning result and the recognition result, converting the receiving signal strength into a pseudo-color image, each receiving signal strength corresponding to a pixel point in the pseudo-color image, determining an image with shadow information based on the pseudo-color image, the image with shadow information being caused by the appearance of the visible light target, performing feature extraction and feature recognition on the image with shadow information, determining the position of each visible light target and the category of each visible light target, using the shadow feature to represent the target, and improving the robustness and resistance of the visible light sensing technology.

Description

TECHNICAL FIELD

[0001] The application belongs to the technical field of industrial Internet of Things, and particularly relates to a visible light target sensing method and device. BACKGROUND

[0002] Industrial Internet of Things is an extended application of traditional Internet of Things in the industrial field, and involves all links and main parts of industrial manufacturing and Internet communication technology. In an actual industrial environment, intelligent robots / devices need to have low-delay, high-reliability and high-precision sensing capabilities to make decisions when communicating. Compared with traditional sensing media, visible light sensing has the advantages of low power consumption, high sensing precision and high security, and image information is not directly transmitted, which not only protects privacy but also avoids information leakage. Therefore, a light emitting diode (LED) gradually becomes a new choice in a visible light sensing system.

[0003] In the prior art, visible light target sensing aims to improve the performance of a certain index, such as improving the identification or positioning performance, and is greatly affected by the signal-to-noise ratio. This may be related to the two-dimensional Gaussian distribution of the spatial radiation of visible light, which causes the non-uniform distribution of the received signal strength, limits the positioning or identification performance, and leads to unstable sensing of the visible light target. SUMMARY

[0004] In view of the above problems, the present application provides a visible light target sensing method and device to improve the robustness and resistance of visible light target sensing. The specific scheme is as follows:

[0005] A visible light target sensing method, applied to a processor of a visible light sensing system, the visible light sensing system further comprising a plurality of receiving ends and a plurality of sending ends, the receiving ends being used to receive signals sent by the sending ends, the signals sent by the sending ends being visible light, and the method comprising:

[0006] determining the received signal strength of each receiving end;

[0007] converting the received signal strength into a pseudo-color image, each received signal strength corresponding to a pixel point in the pseudo-color image;

[0008] based on the pseudo-color image, determining an image with shadow information, the image with shadow information being caused by the appearance of a visible light target;

[0009] performing feature extraction and feature recognition on the image with shadow information to determine the position of each visible light target and the category of each visible light target, so as to sense the visible light target.

[0010] Optionally, the sending end adopts a light emitting diode (LED) to send a signal, the receiving end adopts a photodetector (PD) to receive the signal, and the determination of the received signal strength of each receiving end comprises:

[0011] obtaining a first channel transfer function, an emission power of the LED, and a zero-mean Gaussian white noise, the first channel transfer function being a channel transfer function of all PD sets;

[0012] determining the received signal strength of each receiving end based on the first channel transfer function, the emission power of the LED, and the zero-mean Gaussian white noise.

[0013] Optionally, the obtaining of the first channel transfer function comprises:

[0014] obtaining an incident angle of the PD, a field of view angle of the PD, an area of the PD, a distance between the LED and the PD, a state of a light link, an order of a Lambert model, and a height difference of a preset transceiver, the state of the light link being 0 when the link is blocked and 1 otherwise;

[0015] determining a second channel transfer function based on the order of the Lambert model, the field of view angle of the PD, the area of the PD, the distance between the LED and the PD, the state of the light link, and the height difference of the preset transceiver, the second channel transfer function being a channel transfer function of each PD;

[0016] obtaining the first channel transfer function based on the second channel transfer function.

[0017] Optionally, the pseudo-color image comprises coordinates of the LED and coordinates of the PD; and the determination of the image with shadow information based on the pseudo-color image comprises:

[0018] determining a linear equation based on the coordinates of the LED and the coordinates of the PD, the linear equation being used to indicate a light link;

[0019] judging whether there is an intersection between the linear equation and a preset target expression in a real number domain;

[0020] if there is an intersection between the linear equation and the preset target expression in the real number domain, determining an image with shadow information, the shadow in the image with shadow information being the light link with the intersection, the coordinates of the PD corresponding to the shadow being a shadow node, and an area formed by the shadow node being a shadow area.

[0021] Optionally, the feature extraction and feature recognition of the image with shadow information to determine the positions of the visible light targets and the categories of the visible light targets for the perception of the visible light targets comprise:

[0022] performing multi-scale division on the image with shadow information to obtain a global shadow region and a local shadow region;

[0023] extracting and identifying features of the center of the global shadow region or the local shadow region by using a convolutional neural network model to determine the position of each visible light target and the category of each visible light target, so as to perceive the visible light target.

[0024] Optionally, the training of the convolutional neural network model comprises:

[0025] obtaining a sample image with shadow information;

[0026] using the target position and the target category of the sample image with shadow information as a training label to train the sample image with shadow information, so as to obtain a convolutional neural network model.

[0027] Optionally, after the conversion of the received signal strength into the pseudo-color image, the method further comprises:

[0028] dividing the perception region corresponding to the pseudo-color image into a strong light region and a weak light region.

[0029] A perception device of a visible light target, the device being applied to a processor of a visible light perception system, the visible light perception system further comprising a plurality of receiving ends and a plurality of sending ends, the receiving ends being used to receive signals sent by the sending ends, the signals sent by the sending ends being visible light, and the device comprising:

[0030] a first determination unit configured to determine the received signal strength of each receiving end;

[0031] a conversion unit configured to convert the received signal strength into a pseudo-color image, each received signal strength corresponding to a pixel point in the pseudo-color image;

[0032] a second determination unit configured to determine an image with shadow information based on the pseudo-color image, the image with shadow information being caused by the appearance of a visible light target;

[0033] a third determination unit configured to extract and identify features of the image with shadow information to determine the position of each visible light target and the category of each visible light target, so as to perceive the visible light target.

[0034] Optionally, the sending ends use light emitting diodes (LEDs) to send signals, the receiving ends use photodetectors (PDs) to receive signals, and the first determination unit comprises:

[0035] The first acquisition subunit is configured to acquire a first channel transfer function, an emission power of the LED, and a zero-mean Gaussian white noise, wherein the first channel transfer function is a channel transfer function of all PD sets;

[0036] The first determination subunit is configured to determine a received signal strength of each receiving end based on the first channel transfer function, the emission power of the LED, and the zero-mean Gaussian white noise.

[0037] Optionally, the acquisition subunit comprises:

[0038] The second acquisition subunit is configured to acquire an incident angle of the PD, a field of view angle of the PD, an area of the PD, a distance between the LED and the PD, a state of a light link, an order of a Lambert model, and a height difference of a preset transceiver, wherein the state of the light link is 0 when the link is blocked, and is 1 otherwise;

[0039] The second determination subunit is configured to determine a second channel transfer function based on the order of the Lambert model, the field of view angle of the PD, the area of the PD, the distance between the LED and the PD, the state of the light link, and the height difference of the preset transceiver, wherein the second channel transfer function is a channel transfer function of each PD;

[0040] The third acquisition subunit is configured to acquire the first channel transfer function based on the second channel transfer function.

[0041] Based on the above technical solutions, the method and device for sensing a visible light target provided by the application determine received signal strengths of each receiving end, wherein the transmitted signal is visible light, and the integration of the target is realized through the directional radiation and light intensity consistency of the visible light, the positioning result and the recognition result are mutually assisted, the received signal strengths are converted into a pseudo-color image, each received signal strength corresponds to a pixel point in the pseudo-color image, based on the pseudo-color image, an image with shadow information is determined, the image with shadow information is caused by the appearance of the visible light target, feature extraction and feature recognition are performed on the image with shadow information, the positions of each visible light target and the categories of each visible light target are determined, the shadow feature is used to represent the target, and the robustness and resistance of the visible light sensing technology are improved. BRIEF DESCRIPTION OF DRAWINGS

[0042] Various other advantages and benefits will become apparent to those of ordinary skill in the art upon reading the following detailed description of the preferred embodiments. The accompanying drawings are included to provide a description of preferred embodiments, and are not meant to limit the scope of the present application. Furthermore, the same reference numerals in different drawings represent the same or similar components. In the drawings:

[0043] Figure 1A flowchart of a method for sensing a visible light target according to an embodiment of the present application is disclosed.

[0044] Figure 2 A flowchart of a method for determining the received signal strength of each receiving end according to an embodiment of the present application is disclosed.

[0045] Figure 3 A flowchart of a method for obtaining a first channel transfer function according to an embodiment of the present application is disclosed.

[0046] Figure 4 A flowchart of a method for determining an image with shadow information based on a pseudo-color image according to an embodiment of the present application is disclosed.

[0047] Figure 5 A two-dimensional diagram of a visible light sensing area and a light link according to an embodiment of the present application is disclosed.

[0048] Figure 6 A flowchart of a method for feature extraction and feature recognition on an image with shadow information, determining the position of each visible light target and the category of each visible light target, and sensing the visible light target according to an embodiment of the present application is disclosed.

[0049] Figure 7 A flowchart of a training method of a convolutional neural network model according to an embodiment of the present application is disclosed.

[0050] Figure 8 A sample image of a cube or spherical target with shadow information according to an embodiment of the present application is disclosed.

[0051] Figure 9 A structural diagram of a visible light sensing device according to an embodiment of the present application is disclosed. DETAILED DESCRIPTION

[0052] The technical solutions in the embodiments of the present application will be described clearly and completely below with reference to the accompanying drawings in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, rather than all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative work fall within the scope of protection of the present application.

[0053] The terms "first", "second", and the like in the description and in the claims of the present application and above-mentioned drawings are used to distinguish similar objects, and are not necessarily used to describe a particular sequential or chronological order. It should be understood that the terms thus used can be interchanged under appropriate circumstances, and are merely employed to distinguish one object from another object of the same property in the description of the embodiments of the present application. In addition, the terms "include" and "have" and any variations thereof are intended to cover non-exclusive inclusion, so that a process, a method, a system, a product, or an apparatus including a series of units does not have to be limited to those units, but can include other units not clearly listed or inherent to the process, the method, the product, or the apparatus.

[0054] In order to improve the robustness and resistance of the perception of the visible light target, the present application provides a perception method of a visible light target, which will be further described below in combination with the drawings and specific embodiments.

[0055] Please refer to the accompanying Figure 1 , Figure 1 A flowchart of a perception method of a visible light target provided by an embodiment of the present application. The method is applied to a processor of a visible light perception system, the visible light perception system further includes a plurality of receiving ends and a plurality of sending ends, the receiving ends are used to receive signals sent by the sending ends, the signals sent by the sending ends are visible light, and the method can include the following steps:

[0056] Step S101: determining the received signal strength of each receiving end.

[0057] In the present application, a received signal strength value can be measured at each point in space, and the processor acquires the received signal strength of each receiving end, the received signal strength is non-uniformly distributed.

[0058] Step S102: converting the received signal strength into a pseudo-color image, each received signal strength corresponds to a pixel point in the pseudo-color image.

[0059] In the present application, each received signal strength value corresponds to a pixel point, so the received signal strength can be converted into a pseudo-color image.

[0060] Step S103: determining an image with shadow information based on the pseudo-color image, the image with shadow information is caused by the appearance of the visible light target.

[0061] In the present application, when the target appears, part of the light link will be blocked, thereby causing a significant change in the received signal strength, and the change in the image is the image with shadow information obtained from the pseudo-color image, which will be described in detail in the following embodiments, and will not be described here.

[0062] It should be noted that there are various visible light targets in the industrial Internet of Things scene, such as automatic guided vehicles (AGVs), cargo boxes and other production elements.

[0063] Step S104: feature extraction and feature recognition are performed on the image with shadow information to determine the positions of the various visible light targets and the categories of the various visible light targets, so as to perceive the visible light targets.

[0064] In the present application, a classifier and a bounding box regression are used to perform feature extraction and feature recognition on the image with shadow information to determine the positions of the various visible light targets and the categories of the various visible light targets, which will be described in detail in the following embodiments and will not be described here.

[0065] To sum up, the perception method and device for visible light targets provided by the present application determine the received signal strength of each receiving end, the transmitted signal is visible light, the integrated perception of the target is realized through the directional radiation and light intensity consistency of the visible light, the positioning result and the recognition result are mutually assisted, the received signal strength is converted into a pseudo-color image, each received signal strength corresponds to a pixel point in the pseudo-color image, the image with shadow information is determined based on the pseudo-color image, the image with shadow information is caused by the appearance of the visible light target, feature extraction and feature recognition are performed on the image with shadow information to determine the positions of the various visible light targets and the categories of the various visible light targets, the target is represented by the shadow feature, and the robustness and resistance of the visible light perception technology are improved.

[0066] On the basis of the above-mentioned embodiments disclosed in the present application, in another embodiment of the present application, the specific implementation manner of step S101 of determining the received signal strength of each receiving end is described in detail.

[0067] As an implementable manner, please refer to the accompanying Figure 2 , a flowchart of a method for determining the received signal strength of each receiving end disclosed in the present application. The transmitting end transmits a signal by using a light-emitting diode (LED), and the receiving end receives the signal by using a photodetector (PD). The method can include the following steps:

[0068] Step S201: a first channel transfer function, an emission power of an LED and a zero-mean Gaussian white noise are obtained. The first channel transfer function is a channel transfer function of all PD sets.

[0069] It should be noted that the first channel transfer function is H(t), the LED continuously transmits “0” and “1” signals, the average power in the LED signal period is taken as the emission power, the emission power of the LED is P t , and the zero-mean Gaussian white noise is v(t).

[0070] Step S202: determining the received signal strength of each receiving end based on the first channel transfer function, the emission power of the LED, and the zero-mean Gaussian white noise.

[0071] It should be noted that the received signal strength R(t) in the time t can be defined as: R(t) = H(t)P + v(t), that is, all the LED signal strengths received by all the PDs can be represented as: t t + v(t), that is, all the LED signal strengths received by all the PDs can be represented as:

[0072]

[0073] On the basis of the above-mentioned embodiments of the present application, in another embodiment of the present application, the specific implementation of step S201 of obtaining the first channel transfer function is described in detail.

[0074] As an implementable manner, please refer to the accompanying drawings Figure 3 , which is a flowchart of a method for obtaining a first channel transfer function according to the present application. The method can include the following steps:

[0075] Step S301: obtaining the incident angle of the PD, the field of view angle of the PD, the area of the PD, the distance between the LED and the PD, the state of the light link, the order of the Lambert model, and the height difference of the preset transceiver, the state of the light link being 0 when the link is blocked and 1 otherwise.

[0076] It should be noted that the incident angle of the PD is ψ, the field of view angle of the PD is FOV, the area of the PD is A r , the distance between the LED and the PD is d s,r , the state of the light link is l s,r , which is 0 if the link is blocked and 1 otherwise, and the order of the Lambert model is n. The height difference of the preset transceiver is h.

[0077] Step S302: determining a second channel transfer function based on the order of the Lambert model, the field of view angle of the PD, the area of the PD, the distance between the LED and the PD, the state of the light link, and the height difference of the preset transceiver, the second channel transfer function being the channel transfer function of each PD.

[0078] It should be noted that the second channel transfer function can be represented as:

[0079] wherein rect(·) is a rectangular function.

[0080] Step S303: obtaining the first channel transfer function based on the second channel transfer function. ​

[0081] In the present application, the second channel transfer function is the channel transfer function of each PD, the first channel transfer function is the channel transfer function of all PDs, and H(t) is a matrix vector, which can be expressed as H(t).

[0082] On the basis of the above-mentioned embodiments of the present application, in another embodiment of the present application, the specific implementation of step S103 of determining the image with shadow information based on the pseudo-color image is described in detail.

[0083] As an implementation manner, please refer to the accompanying drawings Figure 4 The present application discloses a flowchart of a method for determining an image with shadow information based on a pseudo-color image. The pseudo-color image includes the coordinates of the LED and the coordinates of the PD, and the method can include the following steps:

[0084] Step S401: determining a linear equation based on the coordinates of the LED and the coordinates of the PD, the linear equation being used to indicate a light link;

[0085] It should be noted that the known coordinates of the LED are The coordinates of the PD Obtain the linear equation between two points in space, which represents the light link, and the light link is the link between the LED and the PD.

[0086] Step S402: determining whether there is an intersection between the linear equation and the preset target expression in the real number field.

[0087] Step S403: if there is an intersection between the linear equation and the preset target expression in the real number field, determining the image with shadow information, in which the shadow is the light link with the intersection, the coordinates of the PD corresponding to the shadow are the shadow nodes, and the area formed by connecting the shadow nodes is the shadow area.

[0088] In order to facilitate understanding, Figure 5 It is a two-dimensional schematic diagram of the visible light sensing area and the light link, in which the LED, the visible light target, the shadow node, the shadow area and the light link are labeled.

[0089] In summary, the method for determining the image with shadow information based on the pseudo-color image provided by the embodiments of the present application characterizes the target by the shadow feature of the visible light target, fully utilizes the ubiquitous lighting devices, and is conducive to the low-cost implementation of the sensing of the visible light target.

[0090] On the basis of the above disclosed embodiments of the present application, in still another embodiment of the present application, the specific implementation manner of step S104 of performing feature extraction and feature recognition on the image with shadow information, determining the position of each visible light target and the category of each visible light target, and perceiving the visible light target is described in detail.

[0091] As an implementable manner, please refer to the accompanying Figure 6 The flowchart of a method for performing feature extraction and feature recognition on the image with shadow information, determining the position of each visible light target and the category of each visible light target, and perceiving the visible light target disclosed by the present application is shown. The method can include the following steps:

[0092] Step S501: performing multi-scale division on the image with shadow information to obtain a global shadow region and a local shadow region.

[0093] In the present application, the multi-scale compensation strategy is used to perform multi-scale division on the image with shadow information to obtain a global shadow region and a local shadow region. Both the global shadow region and the local shadow region can represent the attributes of the target.

[0094] Step S502: performing feature extraction and feature recognition on the center of the global shadow region or the local shadow region by a convolutional neural network model, determining the position of each visible light target and the category of each visible light target, and perceiving the visible light target.

[0095] In the present application, the local shadow region more significantly represents the category attribute at the center of the perception region, while the global shadow region is more resistant to occlusion at the boundary of the perception region. Therefore, the position of the visible light target can be represented by the center of the local shadow region or the global shadow region.

[0096] In summary, the method for performing feature extraction and feature recognition on the image with shadow information, determining the position of each visible light target and the category of each visible light target provided by the embodiments of the present application uses the multi-scale compensation strategy, so that the system can adaptively select the required scale of target category and positioning at any position in the perception region, and has certain robustness and anti-occlusion property.

[0097] On the basis of the above disclosed embodiments of the present application, in still another embodiment of the present application, the specific implementation manner of training the convolutional neural network model is described in detail.

[0098] As an implementable manner, please refer to the accompanying Figure 7 The flowchart of a method for training a convolutional neural network model disclosed by the present application is shown. The method can include the following steps:

[0099] Step S601: Obtain a sample image with shadow information.

[0100] For the convenience of understanding, Figure 8 For the sample image with shadow information of a cube or spherical target, the local shadow area and the global shadow area of the spherical target are labeled in the figure.

[0101] Step S602: Take the target position and target category of the sample image with shadow information as a training label, train the sample image with shadow information, and obtain a convolutional neural network model.

[0102] In the present application, the shape of the modeling target can be spherical (radius r) and located at (x t ,y t ,z t ).

[0103] The shadow node of the target can be given by the following formula:

[0104]

[0105]

[0106] (x-x t ) 2 +(y-y t ) 2 +(z-z t ) 2 =r 2

[0107]

[0108] x, y, z [0, 2r]

[0109] The shape of the modeling target can also be a cube. In order to ensure accuracy and reduce computational complexity, the three surfaces closest to each LED are selected to obtain the shadow node. Given a vertex a0=(x o ,y o ,z o ) of a cube, a=length, b=width, c=height, the distances from the light source to the centers of the six surfaces are calculated, and the formulas of the six surfaces in the cube are as follows:

[0110]

[0111] In summary, the method for training the convolutional neural network model provided in the embodiments of the present application constructs a shadow model for targets of different shapes, and realizes visible light target recognition and positioning from the perspective of the image through the shadow model.

[0112] On the basis of the above-mentioned embodiments of the present application, in another embodiment of the present application, the specific implementation mode after the received signal strength is converted into a pseudo-color image is described in detail.

[0113] As an implementable mode, after the received signal strength is converted into a pseudo-color image,

[0114] The pseudo-color image corresponding perception area can also be divided into a strong light area and a weak light area.

[0115] In the present application, the strong light area PD is sparsely arranged, and the weak light area PD is densely arranged.

[0116] In summary, the embodiments of the present application consider that the illumination distribution around the perception area is obviously uneven, and the shadow information is seriously missing, and propose a block layout mode of the sensing area. Although some shadow areas of the visible light target are lost, the remaining shadows can also be clearly observed.

[0117] The method is described in detail in the above-mentioned embodiments of the present application, and the method of the present application can be implemented in various forms of devices. Therefore, the present application also discloses a visible light target perception device, and specific embodiments are given below for detailed description.

[0118] Please refer to the accompanying Figure 9 , Figure 9 Fig. 1 is a structural schematic diagram of a visible light target perception device disclosed by the embodiments of the present application. The device is applied to a processor of a visible light perception system, and the visible light perception system further includes a plurality of receiving ends and a plurality of sending ends. The receiving end is used to receive the signal sent by the sending end, and the signal sent by the sending end is visible light. The device includes:

[0119] A first determination unit 11 is used to determine the received signal strength of each receiving end.

[0120] A conversion unit 12 is used to convert the received signal strength into a pseudo-color image. Each received signal strength corresponds to a pixel point in the pseudo-color image.

[0121] A second determination unit 13 is used to determine an image with shadow information based on the pseudo-color image. The image with shadow information is caused by the appearance of a visible light target.

[0122] A third determination unit 14 is used to perform feature extraction and feature recognition on the image with shadow information, determine the position of each visible light target and the category of each visible light target, and perceive the visible light target.

[0123] As an implementable manner, the sending end adopts a light emitting diode (LED) to send a signal, the receiving end adopts a photodetector (PD) to receive the signal, and the first determining unit 11 comprises:

[0124] A first obtaining sub-unit is configured to obtain a first channel transfer function, an emission power of the LED, and a zero-mean Gaussian white noise, the first channel transfer function being a channel transfer function of all PD sets.

[0125] A first determining sub-unit is configured to determine a received signal strength of each receiving end based on the first channel transfer function, the emission power of the LED, and the zero-mean Gaussian white noise.

[0126] As an implementable manner, the obtaining sub-unit comprises:

[0127] A second obtaining sub-unit is configured to obtain an incident angle of the PD, a field of view angle of the PD, an area of the PD, a distance between the LED and the PD, a state of a light link, and a height difference of a preset transceiver, the state of the light link being 0 when the link is blocked and 1 otherwise.

[0128] A second determining sub-unit is configured to determine an order of a Lambert model based on the incident angle of the PD.

[0129] A third determining sub-unit is configured to determine a second channel transfer function based on the order of the Lambert model, the field of view angle of the PD, the area of the PD, the distance between the LED and the PD, the state of the light link, and the height difference of the preset transceiver, the second channel transfer function being a channel transfer function of each PD.

[0130] A third obtaining sub-unit is configured to obtain the first channel transfer function based on the second channel transfer function.

[0131] The various embodiments in the specification are described in a progressive manner, and each embodiment focuses on the difference from other embodiments, and the same or similar parts between the various embodiments can be referred to each other. For the apparatus disclosed by the embodiments, since it corresponds to the method disclosed by the embodiments, the description is relatively simple, and the related parts can be referred to the method part.

[0132] In addition, it should be noted that the apparatus embodiments described above are merely exemplary, and the units described as separate components can or can not be physically separate, and the components shown as units can or can not be physical units, i.e., they can be located in one place, or distributed on multiple network units. Some or all of the modules can be selected to achieve the purpose of the embodiment according to actual needs. In addition, the connection between the modules in the apparatus embodiments provided by the present application indicates that there is a communication connection between them, which can be implemented as one or more communication buses or signal lines. Those skilled in the art can understand and implement it without creative labor.

[0133] Through the description of the above embodiments, those skilled in the art can clearly understand that the present application can be realized by means of software and necessary general hardware, and of course can also be realized by special hardware including special integrated circuits, special CPUs, special memories, special components, etc. Generally, functions completed by computer programs can be easily realized by corresponding hardware, and specific hardware structures for realizing the same function can also be various, such as analog circuits, digital circuits or special circuits. However, for the present application, software program implementation is a better embodiment. Based on this understanding, the technical solutions of the present application or the parts that contribute to the prior art can be embodied in the form of software products, which are stored in readable storage media, such as computer floppy disks, U disks, mobile hard disks, read-only memories (ROM), random access memories (RAM), magnetic or optical disks, etc., including a number of instructions to make a computer device (which can be a personal computer, a server, or a network device, etc.) execute the methods described in various embodiments of the present application.

[0134] In summary, the above embodiments are only used to illustrate the technical solutions of the present application, rather than limit them. Although the present application has been described in detail with reference to the above embodiments, those skilled in the art should understand that they can modify the technical solutions recorded in the above embodiments, or make equivalent replacement for some technical features. The modification or replacement does not make the corresponding technical solution deviate from the spirit and scope of the technical solutions of the embodiments of the present application.

Claims

1. A method of perceiving a visible light target, characterized by, The method is applied to a processor of a visible light perception system, the visible light perception system further comprising a plurality of receiving ends and a plurality of sending ends, the receiving ends being configured to receive signals sent by the sending ends, the signals sent by the sending ends being visible light, and the method comprising: determining a receiving signal strength of each receiving end; converting the receiving signal strength into a pseudo-color image, each receiving signal strength corresponding to a pixel point in the pseudo-color image; based on the pseudo-color image, determining an image with shadow information, the image with shadow information being caused by the appearance of a visible light target; performing feature extraction and feature recognition on the image with shadow information to determine the position of each visible light target and the category of each visible light target, so as to perceive the visible light target; the sending end uses a light-emitting diode (LED) to send signals, and the receiving end uses a photodetector (PD) to receive signals, and the determination of the receiving signal strength of each receiving end comprises: obtaining a first channel transfer function, an emission power of the LED, and a zero-mean Gaussian white noise, the first channel transfer function being a channel transfer function of all PDs; based on the first channel transfer function, the emission power of the LED, and the zero-mean Gaussian white noise, determining the receiving signal strength of each receiving end; the obtaining of the first channel transfer function comprises: obtaining an incident angle of the PD, a field of view angle of the PD, an area of the PD, a distance between the LED and the PD, a state of a light link, an order of a Lambert model, and a height difference of a preset transceiver, the state of the light link being 0 when the link is blocked and 1 otherwise; based on the order of the Lambert model, the field of view angle of the PD, the area of the PD, the distance between the LED and the PD, the state of the light link, and the height difference of the preset transceiver, determining a second channel transfer function, the second channel transfer function being a channel transfer function of each PD; based on the second channel transfer function, obtaining the first channel transfer function.

2. The method of claim 1, wherein, The pseudo-color image includes coordinates of the LED and coordinates of the PD; and the determination of the image with shadow information based on the pseudo-color image comprises: based on the coordinates of the LED and the coordinates of the PD, determining a linear equation, the linear equation being used to indicate a light link; judging whether there is an intersection between the linear equation and a preset target expression in a real number domain; if there is an intersection between the linear equation and the preset target expression in the real number domain, determining an image with shadow information, the shadow in the image with shadow information being the light link with the intersection, the coordinates of the PD corresponding to the shadow being a shadow node, and an area formed by the shadow node being a shadow area.

3. The method of claim 1, wherein, The feature extraction and feature recognition on the image with shadow information to determine the position of each visible light target and the category of each visible light target, so as to perceive the visible light target, comprise: performing multi-scale division on the image with shadow information to obtain a global shadow area and a local shadow area; The center of the global shadow area or the local shadow area is subjected to feature extraction and feature recognition by a convolutional neural network model, so as to determine the position of each visible light target and the category of each visible light target, thereby realizing the perception of the visible light target.

4. The method of claim 3, wherein, The training of the convolutional neural network model comprises: obtaining a sample image with shadow information; training the sample image with shadow information by taking the target position and target category of the sample image with shadow information as a training label, to obtain a convolutional neural network model.

5. The method of claim 1, wherein, After the conversion of the received signal strength into the pseudo-color image, the method further comprises: dividing the perception area corresponding to the pseudo-color image into a strong light area and a weak light area.

6. A device for sensing a visible light target, characterized in that The device is applied to a processor of a visible light perception system, the visible light perception system further comprising a plurality of receiving ends and a plurality of sending ends, the receiving ends being configured to receive signals sent by the sending ends, the signals sent by the sending ends being visible light, and the device comprising: a first determination unit configured to determine the received signal strength of each receiving end; a conversion unit configured to convert the received signal strength into a pseudo-color image, each received signal strength corresponding to a pixel point in the pseudo-color image; a second determination unit configured to determine an image with shadow information based on the pseudo-color image, the image with shadow information being caused by the appearance of a visible light target; a third determination unit configured to perform feature extraction and feature recognition on the image with shadow information, to determine the position of each visible light target and the category of each visible light target, thereby realizing the perception of the visible light target; the sending ends send signals by using light-emitting diodes (LEDs), and the receiving ends receive signals by using photodetectors (PDs), and the first determination unit comprises: a first acquisition subunit configured to acquire a first channel transfer function, the emission power of the LED, and zero-mean Gaussian white noise, the first channel transfer function being the channel transfer function of all PDs; a first determination subunit configured to determine the received signal strength of each receiving end based on the first channel transfer function, the emission power of the LED, and the zero-mean Gaussian white noise; the first acquisition subunit comprises: a second acquisition subunit configured to acquire the incident angle of the PD, the field of view angle of the PD, the area of the PD, the distance between the LED and the PD, the state of a light link, the order of a Lambert model, and the height difference of a preset transceiver, the state of the light link being 0 when the link is blocked and 1 otherwise; a second determination subunit configured to determine a second channel transfer function based on the order of the Lambert model, the field of view angle of the PD, the area of the PD, the distance between the LED and the PD, the state of the light link, and the height difference of the preset transceiver, the second channel transfer function being the channel transfer function of each PD; a third acquisition subunit configured to acquire the first channel transfer function based on the second channel transfer function.

Images