Passive reflection type target identification method based on linearly polarized light and identification system thereof

Abstract

The invention provides a passive reflective target identification method based on linearly polarized light. The method comprises the following steps: acquiring a plurality of target images; performing preprocessing operation on the target image to obtain a target feature image; for the target feature image, obtaining a target edge image by means of edge detection and circle fitting; for a target edge image, extraction and coding of the passive reflective target are performed, and an identification system is provided based thereon, comprising a passive reflective target, a strobe incandescent lamp, a color camera and a data processing unit. The target light source is enabled to actively stroboflash through the high-frequency asynchronous signal coding technology, the passive reflection type target is designed based on the optical reflection principle, strong light and electromagnetic interference are avoided, the measurement result is stable and reliable, the miniaturization and passive design requirements are met, and the high practical application value is achieved.

Description

Technical Field

[0001] The invention belongs to the technical field of target recognition in complex environments, and in particular to a passive reflective target recognition method based on linearly polarized light and a recognition system thereof. Background Art

[0002] Under strong light background, target imaging faces many complex and difficult problems, which makes target recognition an extremely challenging task.

[0003] Strong light can cause image sensors to generate a lot of noise. These noises are of various types, including thermal noise, shot noise, etc., which are intertwined with useful target signals. Traditional filtering methods often blur the edges and details of the image while removing noise, making the target features unclear and affecting the accuracy of subsequent recognition. For example, mean filtering may blur the edges of the target in the process of smoothing noise, making it difficult to accurately extract its shape features. At the same time, large areas of clutter may be generated under strong light backgrounds, such as reflected light and light spots. These clutters are similar to the target in terms of color, texture, etc. and are easy to confuse. Edge detection-based methods may misjudge the edges of clutter as target edges, resulting in a large number of false detection areas. For machine learning methods, the presence of clutter increases the complexity of the sample, making it difficult for the model to accurately learn the characteristic pattern of the target, reducing the accuracy and reliability of classification. For example, when using SVM or CNN for training, the clutter area may be mistakenly learned as the feature of the target, resulting in misclassification in actual recognition. Due to the influence of noise and clutter, key features of the target, such as specific shapes, color markings, textures, etc., are difficult to accurately extract. Under strong light, the color of the target may be distorted, and the originally clear texture may be covered by noise, making the recognition method based on color and texture features ineffective; the target may be partially blocked or deformed under strong light, which further increases the difficulty of feature extraction and makes it difficult to find stable and distinguishing features for effective target recognition. Therefore, in order to overcome the shortcomings of the existing technology in noise and clutter processing under strong light background, it is urgent and necessary to seek a passive reflective target recognition method and recognition system based on linear polarized light to achieve accurate recognition of target position and posture. Summary of the invention

[0004] In view of the defects in the above-mentioned prior art, the present invention proposes a passive reflective target recognition method based on linearly polarized light. The method includes collecting a number of target images; performing a preprocessing operation on the target image to obtain a target feature image; obtaining a target edge image by means of edge detection and circle fitting on the target feature image; extracting and encoding the passive reflective target on the target edge image, and based on this, providing an identification system, which includes a passive reflective target, a strobe incandescent lamp, a color camera and a data processing unit. The present invention uses high-frequency asynchronous signal coding technology to make the target light source actively strobe, and designs a passive reflective target based on the principle of optical reflection. It is not affected by strong light and electromagnetic interference, and the measurement results are stable and reliable. It meets the requirements of miniaturization and passive design, and has strong practical application value.

[0005] The present invention provides a passive reflective target recognition method based on linearly polarized light, which comprises the following steps:

[0006] S1. Collecting a number of target images: using a color camera to collect a number of target images of passive reflective targets that are illuminated by a stroboscopic incandescent lamp to cause passive stroboscopic strobing;

[0007] S2, performing a preprocessing operation on the target image to obtain a target feature image;

[0008] S21, using a weighted average method, converting the target image into a target grayscale image through a grayscale conversion formula; the grayscale conversion formula is:

[0009] G=0.299R+0.587G+0.114B (1)

[0010] Wherein, G represents the grayscale pixel value; R, G, and B represent the pixel values ​​of the corresponding positions of the red channel, green channel, and blue channel in the target image, respectively;

[0011] S22, converting the target grayscale image into a target denoised image by means of adaptive median filtering;

[0012] S23, obtaining a target feature image with background removed by using a frame difference method on the target denoised images at adjacent moments;

[0013] S3, obtaining a target edge image by edge detection and circle fitting for the target feature image;

[0014] S31, extracting edge and detail information of the passive reflective target in the target feature image by means of a gradient-based edge detection algorithm to obtain a target edge image;

[0015] S32, determining the center coordinates and radius of the passive reflective target in the target edge image by means of a circle fitting equation;

[0016] S4, extracting and encoding the passive reflective target based on the target edge image;

[0017] S41, based on the target edge image and the target image, obtaining the target image containing only the passive reflective target area and using it as the target image to be encoded;

[0018] S42, identifying the color of the passive reflective target in the target image to be encoded and using it as a first encoding basic element: extracting a first encoding basic element related to the color of the passive reflective target in the target image to be encoded, wherein the first encoding basic element includes a main color tone, a key color share, and a color distribution feature;

[0019] S43, recording and quantizing the time domain information of the passive reflective target in the target image to be encoded as a second encoding basic element, wherein the second encoding basic element includes a time point, a duration, and a position in a time series;

[0020] S44. Based on the first coding basic element and the second coding basic element, according to a preset coding logic, a unique code dedicated to the passive reflective target is generated as a passive reflective target code.

[0021] Further, the step S1 specifically includes the following steps:

[0022] S11, emitting two overlapping and flashing beams of transverse polarized light and longitudinal polarized light through a stroboscopic incandescent lamp, irradiating the surface of the target to be measured, and stimulating the passive reflective target on the surface of the target to be measured to passively strobe;

[0023] S12, using a color camera to photograph the passive reflective target area on the surface of the target to be measured, and collecting a number of target images.

[0024] Preferably, in the step S44, a predefined color code table is used for the main color coding in the first coding basic element, in which red, blue and yellow are encoded as 01, 02 and 03 respectively; the key color proportion coding is quantized into a numerical value in the [0,1] interval and then compressed by Huffman coding; the color distribution characteristics are encoded in the form of a simplified topological graph, the connectivity and position relationship of the color area are marked, and a color code string is generated.

[0025] Preferably, the window size of the adaptive median filter in step S22 is dynamically adjusted based on noise intensity, and the window size ranges from 3×3 to 9×9 pixels.

[0026] Preferably, after the passive reflective target is encoded in step S44, the passive reflective target code is stored in a distributed database in a structured data format, and is transmitted to a monitoring center and associated equipment in real time using a 422 protocol.

[0027] Preferably, the shooting frequency of the color camera in step S1 is consistent and synchronized with the flashing frequency of the stroboscopic incandescent lamp, and the field of view shot by the color camera covers all passive reflective targets.

[0028] Another aspect of the present invention provides an identification system using the aforementioned passive reflective target identification method based on linear polarized light, which includes a passive reflective target, a strobe incandescent lamp, a color camera and a data processing unit; the passive reflective target includes a first red reflective target, a second red reflective target, a first blue reflective target, a second blue reflective target, a first yellow reflective target and a second yellow reflective target; the first red reflective target, the second red reflective target, the first blue reflective target, the second blue reflective target, the first yellow reflective target and the second yellow reflective target are respectively glued to the target to be tested. target surface; two stroboscopic incandescent lamps are provided and flash alternately, a polarizer is provided in front of each of the stroboscopic incandescent lamps and two beams of overlapping and flashing transverse polarized light and longitudinal polarized light are emitted through the polarizer respectively, the transverse polarized light and longitudinal polarized light are irradiated to the target surface to be measured and stimulate the passive reflective target to passively flash; the color camera collects the target image of the area where the passive reflective target is located, and inputs several of the target images into the data processing unit; the data processing unit identifies and encodes the passive reflective target in the target image.

[0029] Preferably, the first red reflective target and the second red reflective target each include a red reflective film and a polarizer; the first blue reflective target and the second blue reflective target each include a blue reflective film and a polarizer; the first yellow reflective target and the second yellow reflective target each include a yellow reflective film and a polarizer; the polarization directions corresponding to the first red reflective target and the second red reflective target are perpendicular to each other, the polarization directions corresponding to the first blue reflective target and the second blue reflective target are perpendicular to each other, and the polarization directions corresponding to the first yellow reflective target and the second yellow reflective target are perpendicular to each other.

[0030] Preferably, the first red reflective target, the second red reflective target, the first blue reflective target, the second blue reflective target, the first yellow reflective target and the second yellow reflective target are all circular.

[0031] Preferably, the red reflective film, the blue reflective film and the yellow reflective film are all made of flexible reflective material, and the daytime visible light reflectivity and the nighttime visible light reflectivity are respectively greater than or equal to 60% and 65%.

[0032] Compared with the prior art, the technical effects of the present invention are:

[0033] 1. The passive reflective target recognition method based on linearly polarized light designed by the present invention allows the target light source to actively flash through high-frequency asynchronous signal coding technology, and uses the frame difference method to separate and identify the target image according to the background, target and noise characteristics. The passive reflective target is designed based on the optical reflection principle, and the light is reflected by a special reflective surface to achieve accurate recognition of the target position and posture, which is more practical.

[0034] 2. The passive reflective target recognition method based on linearly polarized light designed by the present invention does not need to modulate the energy output process when the environment is a complex strong light interference environment in the air, which reduces the complexity and cost of the system and is not affected by factors such as electromagnetic interference. It is suitable for a variety of application scenarios including severe weather such as atmospheric turbulence or rain and fog. The measurement results are stable and reliable, meeting the requirements of miniaturization and passive design. BRIEF DESCRIPTION OF THE DRAWINGS

[0035] Other features, objects and advantages of the present application will become more apparent from the detailed description of non-limiting embodiments made with reference to the following drawings.

[0036] Figure 1 It is a flow chart of the passive reflective target recognition method based on linearly polarized light of the present invention;

[0037] Figure 2 It is a schematic diagram of the polarized light reflection principle of the present invention;

[0038] Figure 3 is a schematic diagram of target image edge extraction in a specific embodiment of the present invention;

[0039] Figure 4 It is a schematic diagram of passive reflective target arrangement in a specific embodiment of the present invention. DETAILED DESCRIPTION

[0040] The present application will be further described in detail below in conjunction with the accompanying drawings and embodiments. It will be appreciated that the specific embodiments described herein are only used to explain the relevant inventions, rather than to limit the inventions. It should also be noted that, for ease of description, only portions related to the relevant inventions are shown in the accompanying drawings. It should be noted that, in the absence of conflict, the embodiments in the present application and the features in the embodiments may be combined with each other. The present application will be described in detail below with reference to the accompanying drawings and in conjunction with the embodiments.

[0041] Figure 1 The passive reflective target recognition method based on linearly polarized light of the present invention is shown, and the method comprises the following steps:

[0042] S1. Collecting a number of target images: using a color camera to collect a number of target images of passive reflective targets that are illuminated by stroboscopic incandescent lamps to cause passive strobing.

[0043] S11, two overlapping and flashing beams of transverse polarized light and longitudinal polarized light are emitted by the stroboscopic incandescent lamp, and irradiated onto the surface of the target to be measured, thereby stimulating the passive reflective target on the surface of the target to be measured to passively flash; Figure 2 As shown, polarized light is irradiated onto the reflective film surface of the target and reflected back, and the target passively flashes at a specific frequency, while the background stray light is constantly bright or changes irregularly, which is used to cut out the noise.

[0044] S12, using a color camera to photograph the passive reflective target area on the surface of the target to be measured, and collecting a number of target images. The frequency of the color camera photographing is consistent and synchronized with the flashing frequency of the strobe incandescent lamp, and the field of view of the color camera photographing covers all passive reflective targets.

[0045] Use high-resolution, wide-dynamic-range optical imaging equipment, such as industrial-grade CCD high-speed cameras, to capture images of scenes containing passive reflective targets from a specific observation angle to ensure that the target area is fully included in the field of view. The acquisition frame rate is flexibly set based on the dynamic characteristics of the target and the real-time requirements of the application scenario, generally between 100 and 200 frames per second.

[0046] S2. Perform preprocessing operations on the target image to obtain a target feature image.

[0047] S21. By means of the weighted average method, the target image is converted into a target grayscale image through the grayscale conversion formula, so as to reduce the data dimension and improve the subsequent processing efficiency; the grayscale conversion formula is:

[0048] G=0.299R+0.587G+0.114B (1)

[0049] Among them, G represents the grayscale pixel value; R, G, and B represent the pixel values ​​of the corresponding positions of the red channel, green channel, and blue channel in the target image, respectively.

[0050] S22. With the help of adaptive median filtering, the target grayscale image is converted into a target denoised image, accurately retaining the target edge and detail information and avoiding over-smoothing; the window size of the adaptive median filter is dynamically adjusted based on the noise intensity, and the window size ranges from 3×3 to 9×9 pixels.

[0051] S23, obtaining a target feature image with the background removed by using a frame difference method on the target denoised images at adjacent moments.

[0052] S3, such as Figure 3 As shown, for the target feature image, the target edge image is obtained by edge detection and circle fitting.

[0053] S31. With the help of a gradient-based edge detection algorithm, the edge and detail information of the passive reflective target in the target feature image is extracted to obtain a target edge image with an error controlled within ±0.1 mm.

[0054] S32. Determine the center coordinates and radius of the passive reflective target in the target edge image using a circle fitting equation.

[0055] S4. Extract and encode the passive reflective target based on the target edge image.

[0056] S41. Based on the target edge image and the target image, a target image containing only the passive reflective target area is obtained and used as the target image to be encoded.

[0057] S42, identifying the color of the passive reflective target in the target image to be encoded and using it as the first encoding basic element: extracting the first encoding basic element related to the color of the passive reflective target in the target image to be encoded, the first encoding basic element including main color tone, key color proportion and color distribution characteristics.

[0058] S43, recording the time domain information of the passive reflective target in the target image to be encoded and quantizing it as the second encoding basic element, the second encoding basic element includes the time point, the duration and the position in the time series.

[0059] S44. Based on the first coding basic element and the second coding basic element, according to a preset coding logic, a unique code dedicated to the passive reflective target is generated as the passive reflective target code.

[0060] A predefined color code table is used to encode the main color tone in the first coding basic element. In the predefined color code table, red, blue and yellow are encoded as 01, 02 and 03 respectively; the key color proportion is quantized into a value in the [0,1] interval and then compressed by Huffman coding; the color distribution characteristics are encoded in the form of a simplified topological graph, the connectivity and position relationship of the color area are marked, and a color code string is generated.

[0061] After the passive reflective target is encoded, the passive reflective target code is stored in a distributed database in a structured data format to ensure high data availability and scalability, and is transmitted to the monitoring center and related equipment in real time using the 422 protocol. The transmission delay is controlled within 20ms, meeting application scenarios with demanding real-time requirements.

[0062] Another aspect of the present invention provides an identification system using the aforementioned passive reflective target identification method based on linear polarized light, which includes a passive reflective target, a strobe incandescent lamp, a color camera and a data processing unit; the passive reflective target includes a first red reflective target, a second red reflective target, a first blue reflective target, a second blue reflective target, a first yellow reflective target and a second yellow reflective target; the first red reflective target, the second red reflective target, the first blue reflective target, the second blue reflective target, the first yellow reflective target and the second yellow reflective target are respectively glued to the surface of the target to be measured, such as Figure 4 As shown; the first red reflective target and the second red reflective target each include a red reflective film and a polarizer; the first blue reflective target and the second blue reflective target each include a blue reflective film and a polarizer; the first yellow reflective target and the second yellow reflective target each include a yellow reflective film and a polarizer; the polarization directions corresponding to the first red reflective target and the second red reflective target are perpendicular to each other, the polarization directions corresponding to the first blue reflective target and the second blue reflective target are perpendicular to each other, and the polarization directions corresponding to the first yellow reflective target and the second yellow reflective target are perpendicular to each other.

[0063] The first red reflective target, the second red reflective target, the first blue reflective target, the second blue reflective target, the first yellow reflective target and the second yellow reflective target are all circular; the red reflective film, the blue reflective film and the yellow reflective film are all made of flexible reflective materials and the daytime visible light reflectivity and the nighttime visible light reflectivity are greater than or equal to 60% and 65% respectively.

[0064] In a specific embodiment, the target mainly adopts an integrated design of flexible reflective material and polarizer, the cumulative thickness of the components is no more than 1mm, the size is 25mm in diameter, and high-transmittance horizontal and vertical polarizers and red, yellow and blue diamond-grade reflective films are used.

[0065] Two stroboscopic incandescent lamps are provided and flash alternately, a polarizer is provided in front of each stroboscopic incandescent lamp and two beams of overlapping and flashing transverse polarized light and longitudinal polarized light are emitted through the polarizer respectively, the transverse polarized light and the longitudinal polarized light are irradiated to the surface of the target to be measured and stimulate the passive reflective target to passively flash; the color camera collects the target image of the area where the passive reflective target is located, and inputs several target images into the data processing unit; the data processing unit identifies and encodes the passive reflective target in the target image.

[0066] If the polarizing plate is in the same direction as the polarized light of the stroboscopic incandescent lamp, the target will be bright. If it is perpendicular to the polarized light of the stroboscopic incandescent lamp, the light will not be able to pass through and it will become dark. At the same time, three red, blue and yellow targets will be bright, and three red, blue and yellow targets will be dark. The target numbers are distinguished by the three primary colors. Define time 0, targets ①③⑤ are bright, and targets ②④⑥ are dark; define time 1, targets ①③⑤ are dark, and targets ②④⑥ are bright. As shown in Table 1, the first digit in the code indicates bright (0) or dark (1), and the last two digits indicate the color, where 00 is red, 01 is blue, and 10 is yellow.

[0067]

[0068]

[0069] Table 1

[0070] The invention designs a passive reflective target recognition method based on linearly polarized light. The target light source is actively flashed by high-frequency asynchronous signal coding technology. The target image is separated and recognized according to the characteristics of the background, target and noise by using the frame difference method. The passive reflective target is designed based on the optical reflection principle. The light is reflected by a special reflection surface to realize accurate recognition of the target position and posture, which is more practical. When the environment is a complex strong light interference environment in the air, the proposed method does not need to modulate the energy output process, which reduces the complexity and cost of the system and is not affected by factors such as electromagnetic interference. The method is suitable for a variety of application scenarios including severe weather such as atmospheric turbulence or rain and fog. The measurement results are stable and reliable, meeting the requirements of miniaturization and passive design.

[0071] Finally, it should be noted that the above embodiments are only intended to illustrate rather than limit the technical solutions of the present invention. Although the present invention has been described in detail with reference to the above embodiments, those skilled in the art should understand that the present invention can still be modified or replaced by equivalents. Any modification or partial replacement that does not depart from the spirit and scope of the present invention should be included in the scope of the claims of the present invention.

Claims

1. A passive reflective target recognition method based on linearly polarized light, characterized in that: It includes the following steps: S1. Collecting a number of target images: using a color camera to collect a number of target images of passive reflective targets that are illuminated by a stroboscopic incandescent lamp to cause passive stroboscopic strobing; S2, performing a preprocessing operation on the target image to obtain a target feature image; S21, using a weighted average method, converting the target image into a target grayscale image through a grayscale conversion formula; the grayscale conversion formula is: G=0.299R+0.587G+0.114B (1) Wherein, G represents the grayscale pixel value; R, G, and B represent the pixel values ​​of the corresponding positions of the red channel, green channel, and blue channel in the target image, respectively; S22, converting the target grayscale image into a target denoised image by means of adaptive median filtering; S23, obtaining a target feature image with background removed by using a frame difference method on the target denoised images at adjacent moments; S3, obtaining a target edge image by edge detection and circle fitting for the target feature image; S31, extracting edge and detail information of the passive reflective target in the target feature image by means of a gradient-based edge detection algorithm to obtain a target edge image; S32, determining the center coordinates and radius of the passive reflective target in the target edge image by means of a circle fitting equation; S4, extracting and encoding the passive reflective target based on the target edge image; S41, based on the target edge image and the target image, acquiring the target image containing only the passive reflective target area and using it as the target image to be encoded; S42, identifying the color of the passive reflective target in the target image to be encoded and using it as a first encoding basic element: extracting a first encoding basic element related to the color of the passive reflective target in the target image to be encoded, wherein the first encoding basic element includes a main color tone, a key color share, and a color distribution feature; S43, recording and quantizing the time domain information of the passive reflective target in the target image to be encoded as a second encoding basic element, wherein the second encoding basic element includes a time point, a duration, and a position in a time series; S44. Based on the first coding basic element and the second coding basic element, according to a preset coding logic, a unique code dedicated to the passive reflective target is generated as a passive reflective target code.

2. The passive reflective target recognition method based on linearly polarized light according to claim 1, characterized in that: The step S1 specifically includes the following steps: S11, emitting two overlapping and flashing beams of transverse polarized light and longitudinal polarized light through a stroboscopic incandescent lamp, irradiating the surface of the target to be measured, and stimulating the passive reflective target on the surface of the target to be measured to passively flash; S12, using a color camera to photograph the passive reflective target area on the surface of the target to be measured, and collecting a number of target images.

3. The passive reflective target recognition method based on linearly polarized light according to claim 1, characterized in that: In step S44, a predefined color code table is used for encoding the main color tone in the first coding basic element, in which red, blue and yellow are encoded as 01, 02 and 03 respectively; the key color proportion encoding is quantized into a numerical value in the [0,1] interval and then compressed by Huffman encoding; the color distribution feature is encoded in the form of a simplified topological graph, marking the connectivity and position relationship of the color area to generate a color code string.

4. The passive reflective target recognition method based on linearly polarized light according to claim 1, characterized in that: The window size of the adaptive median filter in step S22 is dynamically adjusted based on the noise intensity, and the window size ranges from 3×3 to 9×9 pixels.

5. The passive reflective target recognition method based on linearly polarized light according to claim 1, characterized in that: After the passive reflective target is encoded in step S44, the passive reflective target code is stored in a distributed database in a structured data format, and is transmitted to the monitoring center and associated equipment in real time using the 422 protocol.

6. The passive reflective target recognition method based on linearly polarized light according to claim 1, characterized in that: In step S1, the shooting frequency of the color camera is consistent and synchronized with the flashing frequency of the stroboscopic incandescent lamp, and the field of view of the color camera covers all passive reflective targets.

7. An identification system for the passive reflective target identification method based on linearly polarized light according to any one of claims 1 to 6, characterized in that: It includes a passive reflective target, a stroboscopic incandescent lamp, a color camera and a data processing unit; the passive reflective target includes a first red reflective target, a second red reflective target, a first blue reflective target, a second blue reflective target, a first yellow reflective target and a second yellow reflective target; the first red reflective target, the second red reflective target, the first blue reflective target, the second blue reflective target, the first yellow reflective target and the second yellow reflective target are respectively glued to the surface of the target to be measured; the stroboscopic incandescent lamp is provided with two and flashes alternately A polarizer is arranged in front of each of the stroboscopic incandescent lamps, and two beams of overlapping and flashing transverse polarized light and longitudinal polarized light are emitted through the polarizer respectively, and the transverse polarized light and longitudinal polarized light are irradiated onto the surface of the target to be measured and stimulate the passive reflective target to passively strobe; the color camera collects the target image of the area where the passive reflective target is located, and inputs several of the target images into the data processing unit; the data processing unit identifies and encodes the passive reflective target in the target image.

8. The identification system according to claim 7, characterized in that: The first red reflective target and the second red reflective target each include a red reflective film and a polarizer; the first blue reflective target and the second blue reflective target each include a blue reflective film and a polarizer; the first yellow reflective target and the second yellow reflective target each include a yellow reflective film and a polarizer; the polarization directions corresponding to the first red reflective target and the second red reflective target are perpendicular to each other, the polarization directions corresponding to the first blue reflective target and the second blue reflective target are perpendicular to each other, and the polarization directions corresponding to the first yellow reflective target and the second yellow reflective target are perpendicular to each other.

9. The identification system according to claim 7 or 8, characterized in that: The first red reflective target, the second red reflective target, the first blue reflective target, the second blue reflective target, the first yellow reflective target and the second yellow reflective target are all circular.

10. The identification system according to claim 9, characterized in that: The red reflective film, the blue reflective film and the yellow reflective film are all made of flexible reflective material, and the daytime visible light reflectivity and the nighttime visible light reflectivity are respectively greater than or equal to 60% and 65%.

Images