A range-gated lidar image reconstruction method

Abstract

The application provides a range-gated laser radar image reconstruction method, which solves the technical problems of existing range-gated laser radar imaging, such as target edge blur, serious environmental noise interference and poor distance resolution. It includes obtaining a plurality of radar slice images, after preprocessing, each frame is sequentially taken as a target frame preprocessed image, difference is performed with adjacent frames before and after, binarization is performed respectively, and operation is performed to obtain a target frame distance information enhancement map; the target frame distance information enhancement map and the target frame preprocessed image are operated to obtain an optimized target frame radar slice image, multiplied by the corresponding detection distance to obtain a target frame distance matrix, and after iteration and superposition, a detection target object matrix is obtained, and pseudo-color processing is performed to reconstruct a detection target object distance image. The application can be widely applied to the field of laser radar imaging technology.

Description

TECHNICAL FIELD

[0001] The application belongs to the technical field of laser radar imaging, and more particularly to a range-gated laser radar image reconstruction method. BACKGROUND

[0002] Laser radar is a technology that uses pulsed laser as a transmitting source, receives the reflected echo signal intensity of the laser beam on the target object, and obtains the distance between the target object and the laser to realize the detection of the shape of the target object according to the time-of-flight method. The characteristics of short wavelength, narrow pulse width, smaller divergence angle and higher pulse energy of laser are used to realize the detection of target objects at a farther distance with higher resolution.

[0003] The range-gated technology realizes the reception of echo signals of specific distances by precisely controlling the transmission time of pulsed laser and the opening time of the gated detector, which is an important method to suppress the backscattering effect of laser. However, when detecting a large target object, the sampling area needs to cover the entire target object. In order to not increase the opening time of the gating, the time width of the range-gated is generally much larger than the delay step length of the multiple continuous step slice images of the detector, resulting in a time overlap between each slice image and the previous and next several images, which increases the background noise of the target object.

[0004] The existing range-gated laser radar image reconstruction technology mostly uses the centroid algorithm and the binary algorithm to separate the target object from the background and obtain the distance characteristics of the target object. However, the binary algorithm is only suitable for images with high contrast between the target object and the background, and the threshold needs to be determined according to the specific situation of each slice image, which is difficult to be used for fast processing of multiple slice images to realize the reconstruction of the distance image of the detected target object. The centroid method depends on the intensity of the target echo signal, and the target object and the background noise are not easy to distinguish, the distance resolution is limited, and the clarity of the reconstructed image is low. SUMMARY

[0005] The purpose of the embodiments of the present application is to provide a range-gated laser radar image reconstruction method to solve the technical problems of blurred edges of the imaging target, serious environmental noise interference and poor distance resolution in the prior art.

[0006] To achieve the above-mentioned purpose, the embodiments of the present application provide a range-gated laser radar image reconstruction method, comprising the following steps:

[0007] A plurality of radar slice images at different distances are obtained, and after preprocessing, each frame is sequentially taken as a target frame preprocessed image starting from the second frame, and difference is performed with the adjacent frames before and after to obtain a target frame preprocessed image and a target frame post-difference image, and after binaryzation, the target frame distance information enhancement image is obtained by performing AND operation.

[0008] The target frame distance information reinforced image and the target frame preprocessed image are subjected to AND operation to obtain an optimized target frame radar slice image, and the distance corresponding to the radar slice image is multiplied to obtain a target frame distance matrix. Iteration is performed until all target frame distance matrices are obtained, and a detection target object matrix is obtained after superposition. The detection target object distance image is reconstructed by pseudo-color processing.

[0009] Preferably, the binarization process respectively includes: binarizing the target frame pre-difference image and the target frame post-difference image respectively with zero as a threshold to obtain a target frame binarized pre-difference image and a target frame binarized post-difference image.

[0010] Preferably, for each pixel intensity value of the pixel intensity matrix of the optimized target frame radar slice image, the pixel intensity value is multiplied by the distance corresponding to the radar slice image to obtain a target frame distance matrix.

[0011] Preferably, the formula for obtaining the optimized target frame radar slice image is:

[0012] ;

[0013] In the formula, is the optimized target frame radar slice image, is the target frame distance information reinforced image, is the target frame preprocessed image.

[0014] Preferably, the formula for obtaining the target frame distance matrix is:

[0015] ;

[0016] In the formula, is the target frame distance matrix, is the optimized target frame radar slice image, is the distance corresponding to the i-th frame radar slice image.

[0017] Preferably, the formula for obtaining the detection target object matrix is:

[0018] ;

[0019] In the formula, is the detection target object matrix, and N is the number of frames of the radar slice image, is the target frame distance matrix.

[0020] Preferably, the target frame distance information reinforced image is obtained by performing AND operation on the target frame binarized pre-difference image and the target frame binarized post-difference image, and the formula is:

[0021] ;

[0022] wherein, is a target frame distance information enhancement map, is a target frame difference image before binarization, is a target frame difference image after binarization.

[0023] Preferably, the preprocessing means: denoising a plurality of radar slice images to obtain a denoised radar slice image.

[0024] Preferably, the denoising is median filtering, mean filtering or Gaussian filtering.

[0025] Preferably, the pseudo-color processing is JET color mapping, piecewise linear mapping or rainbow mapping.

[0026] The application has the beneficial effects that: the application provides a range-gated laser radar image reconstruction method, a plurality of radar slice images of different distances are obtained through continuous stepping of a range-gated gate in a range-gated laser radar, background noise influence caused by the range-gated gate being wider than the laser radar delay step length is overcome, and picture quality is improved; target distance information is strengthened through inter-frame difference and binarization operations to obtain a target frame distance information enhancement map, and background noise is effectively suppressed; an optimized target frame radar slice image is obtained through AND operation of the target frame distance information enhancement map and a target frame preprocessed image, real target edges are highlighted, and blurring is reduced; then, a target frame distance matrix is obtained by multiplying the optimized target frame radar slice image and the corresponding distance of the radar slice image; finally, an iterative calculation target frame distance matrix is obtained, superimposed, a detection target object matrix is obtained, and a high-precision detection target object distance image is reconstructed after pseudo-color processing, and distance resolution and anti-interference performance are significantly improved.

[0027] In summary, the application significantly improves anti-noise ability, optimizes distance resolution performance, and effectively enhances the clarity and authenticity of target edges. BRIEF DESCRIPTION OF DRAWINGS

[0028] In order to more clearly illustrate the technical solutions in the embodiments of the application, the following will briefly introduce the drawings needed to be used in the embodiments or prior art description. Obviously, the drawings in the following description are only some embodiments of the application, and other drawings can be obtained by those skilled in the art without creative labor.

[0029] Figure 1 The figure is a schematic diagram of the overall process of a range-gated laser radar image reconstruction method provided by an embodiment of the application;

[0030] Figure 2In the figure, (a) is the pre-processed image of the target frame, (b) is the pre-processed image of the target frame, and (c) is the post-processed image of the target frame.

[0031] Figure 3 In the figure, (a) is the pre-processed image of the target frame, (b) is the pre-processed image of the target frame, and (c) is the post-processed image of the target frame.

[0032] Figure 4 The optimized target frame radar slice image provided by an embodiment of the present application;

[0033] Figure 5 The target object distance image provided by an embodiment of the present application;

[0034] Figure 6 The contrast figure for detecting an indoor 8-17m scene using the present application, wherein (a) is a schematic diagram of a camera shooting an indoor scene, and (b) is a distance-gated laser radar distance image of the indoor scene.

[0035] Figure 7 The contrast figure for detecting a 20cm interval target at a distance of 10.9m in a water pipe using the present application, wherein (a) is a schematic diagram of a camera shooting an interval target in water, and (b) is a laser radar distance image of the interval target. DETAILED DESCRIPTION

[0036] In order to make the technical problems, technical solutions and beneficial effects of the present application clearer, the present application will be further described in detail below in combination with the drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application and do not limit the present application.

[0037] Please refer to Figure 1 The distance-gated laser radar image reconstruction method provided by an embodiment of the present application comprises:

[0038] S1: Obtain a plurality of radar slice images at different distances, and after preprocessing, sequentially take each frame as a target frame pre-processed image starting from the second frame.

[0039] The gating distance, step number and step length of the distance-gated laser radar are set, N frames of radar slice images at different distances are obtained by continuous stepping, and the obtained N frames of radar slice images at different distances are pre-processed, and sequentially take each frame as a target frame pre-processed image starting from the second frame.

[0040] In an optional embodiment, after obtaining N radar slice images, the present application performs preprocessing to obtain N preprocessed radar slice images, selects three consecutive preprocessed radar slice images (i.e., the i-1th, ith and i+1th frames), takes the i-1th preprocessed radar slice image as a target frame front preprocessed image, as shown in (a) of Figure 2 , takes the ith preprocessed radar slice image as a target frame preprocessed image, as shown in (b) of Figure 2 , and takes the i+1th preprocessed radar slice image as a target frame rear preprocessed image, as shown in (c) of Figure 2 . In the present application, the value of i ranges from 1 to N-1. Specifically, the preprocessing in the present application refers to denoising processing, and the N radar slice images obtained are denoised by using a median filtering method.

[0041] It is worth noting that the method of denoising in the present application is not limited, and any denoising method such as deep learning, graph neural network, median filtering, mean filtering, Gaussian filtering, etc. can be used for denoising.

[0042] S2: Perform difference between the target frame preprocessed image and the adjacent frames to obtain a target frame front difference image and a target frame rear difference image, perform binaryzation processing to obtain a target frame binaryzation front difference image and a target frame binaryzation rear difference image, and perform AND operation to obtain a target frame distance information enhancement image.

[0043] Perform difference between the target frame preprocessed image and the target frame front preprocessed image to obtain the target frame front difference image , and the specific formula is as follows:

[0044] ;

[0045] In the formula, is the target frame front difference image, is the target frame preprocessed image, and is the target frame front preprocessed image.

[0046] Then, perform binaryzation processing on the target frame front difference image with zero as the threshold to obtain the target frame binaryzation front difference image , and the result is shown in (a) of Figure 3 , and the specific formula is as follows:

[0047] ;

[0048] In the formula, is the target frame binaryzation front difference image, and is the target frame front difference image.

[0049] Next, the target frame pre-processing image and the target frame post-processing image are subjected to difference, and the target frame post-difference image is obtained. The specific formula is as follows:

[0050] ;

[0051] In the formula, the target frame post-difference image, is the target frame pre-processing image, and the target frame post-processing image.

[0052] Then, the target frame post-difference image is binarized with zero as the threshold value, and the target frame binarized post-difference image is obtained. The result is shown in (b) of FIG. 6, and the specific formula is as follows: Figure 3

[0053] ;

[0054] In the formula, the target frame binarized post-difference image, is the target frame post-difference image.

[0055] Finally, the target frame binarized pre-difference image and the target frame binarized post-difference image are subjected to logical AND operation, and the target frame distance information reinforced image is obtained. The final result is shown in (c) of FIG. 6, and the formula is as follows: Figure 3

[0056] ;

[0057] In the formula, the target frame distance information reinforced image, is the target frame binarized pre-difference image, and the target frame binarized post-difference image.

[0058] S3: The target frame distance information reinforced image is subjected to AND operation with the target frame pre-processing image to obtain an optimized target frame radar slice image. The distance corresponding to the different distance radar slice image is multiplied to obtain a target frame distance matrix. Iteration is performed until all target frame distance matrices are obtained. Superposition is performed to obtain a detected target object matrix. Pseudo-color processing is performed to reconstruct a detected target object distance image.

[0059] The target frame distance information reinforced image is subjected to AND operation with the target frame pre-processing image ​​The execution logic is an AND operation, and the optimized target frame radar slice image is obtained , as shown in Figure 4 , and the formula is as follows:

[0060] ;

[0061] In the formula, is the optimized target frame radar slice image, is the target frame distance information enhancement map, is the target frame preprocessed image.

[0062] For each pixel intensity value in the pixel intensity matrix of the optimized target frame radar slice image , the distance corresponding to the different distance radar slice image is multiplied, and the target frame distance matrix is obtained, and the formula is as follows:

[0063] ;

[0064] In the formula, is the target frame distance matrix, is the optimized target frame radar slice image, is the distance corresponding to the i-th frame radar slice image.

[0065] The target frame preprocessed image is iterated for 1<i<N until the N-th target frame preprocessed image, so as to obtain the target frame distance matrix corresponding to each frame radar slice image .

[0066] Then, all the obtained target frame distance matrices are superimposed to obtain a detection target object matrix . The specific formula is as follows:

[0067] ;

[0068] In the formula, is the detection target object matrix, and N is the number of frames of the radar slice image, is the target frame distance matrix.

[0069] Then, the detection target object matrix is subjected to pseudo-color processing, and the finally obtained result is as shown in Figure 5 , and the detection target object distance image is reconstructed. The pseudo-color processing can be selected from JET color mapping, piecewise linear mapping, rainbow mapping, etc., and the selection of the pseudo-color processing is not limited here, and can be selected according to actual conditions.

[0070] Specific embodiment 1: imaging experiment of one indoor scene.​

[0071] Referring to Figure 6 , an experimental schematic diagram for detecting an indoor 8-17 m scene by using the distance-gated laser radar image reconstruction method of the present application, wherein, Figure 6 In (a), a camera is used to capture an indoor scene. During the detection process, a distance-gated laser radar is used to continuously step imaging the scene between 8-17 m, with a step time interval of 0.25 ns, corresponding to a step length of 3.75 cm. 240 radar slice images of the detected target objects with equal intervals are obtained. The radar slice images of the detected target objects are reconstructed using the present application, and pseudo-color processing is performed, and finally the distance image of the distance-gated laser radar for detecting the indoor scene is obtained, as shown in Figure 6 (b). According to the color characteristics of the distance image, the distance information between the detected target objects and the laser radar can be accurately obtained.

[0072] Specific embodiment 2: imaging experiment of an underwater interval target.

[0073] Referring to Figure 7 , an experimental schematic diagram for detecting a 20 cm interval target in water at a distance of 10.9 m by using the distance-gated laser radar image reconstruction method of the present application, wherein, Figure 7 In (a), a camera is used to capture a 20 cm interval target in water. During the detection process, a distance-gated laser radar is used to continuously step imaging the range of 9-13 m in water, with a step time interval of 0.25 ns, corresponding to a step length of 2.81 cm in water. 142 frames of slice images of the detected target objects with equal intervals are obtained.

[0074] Due to the strong backscattering of plankton and suspended particles in water, the radar slice images have strong environmental noise. The target object slice images are reconstructed using the present application, and pseudo-color processing is performed, and finally the distance image of the distance-gated laser radar for detecting the interval target in water is obtained, as shown in Figure 7 (b). According to the color characteristics of the distance image, the distance information between the detected interval target and the laser radar can be accurately obtained, with high distance resolution and lateral resolution.

[0075] Those skilled in the art can realize that the units and algorithm steps of each example described in combination with the embodiments disclosed herein can be realized by electronic hardware or a combination of computer software and electronic hardware. Whether the functions are realized in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to realize the described functions for each specific application, but such implementation should not be considered beyond the scope of the present application.

[0076] The above-described embodiments are only used to illustrate the technical solutions of the present application, but not limit them; although the present application is described in detail with reference to the foregoing embodiments, those skilled in the art should understand that: it can still modify the technical solutions recorded in the foregoing embodiments, or make equivalent replacement to part of the technical features; and these modifications or replacements do not make the essence of the corresponding technical solutions deviate from the spirit and scope of the technical solutions of the embodiments of the present application, and should be included in the protection scope of the present application.

Claims

1. A method for reconstructing images using a range-gated lidar system, characterized in that, Includes the following steps: Several radar slice images at different distances are acquired. After preprocessing, starting from the second frame, each frame is used as the target frame preprocessed image. The difference between the target frame and the adjacent frames is performed to obtain the target frame front difference image and the target frame back difference image. After binarization, the AND operation is performed to obtain the target frame distance information enhancement map. The target frame range information enhancement image is ANDed with the target frame preprocessed image to obtain an optimized target frame radar slice image. The target frame range matrix is ​​obtained by multiplying the range image with the range corresponding to the radar slice image. This process is iterated until all target frame range matrices are obtained. The matrices are then superimposed to obtain the target object matrix. Pseudo-color processing is then performed to reconstruct the target object range image.

2. The range-gated lidar image reconstruction method as described in claim 1, characterized in that, The binarization process includes: performing binarization processing on the difference image before the target frame and the difference image after the target frame with zero as the threshold, respectively, to obtain the difference image before binarization and the difference image after binarization of the target frame.

3. The range-gated lidar image reconstruction method as described in claim 1, characterized in that, For each pixel intensity value in the pixel intensity matrix of the optimized target frame radar slice image, the pixel intensity value is multiplied by the distance corresponding to the radar slice image to obtain the target frame distance matrix.

4. The range-gated lidar image reconstruction method as described in claim 1, characterized in that, The formula for obtaining the optimized target frame radar slice image is as follows: ； In the formula, The optimized target frame radar slice image, Enhanced map of target frame distance information. Preprocess the image for the target frame.

5. The range-gated lidar image reconstruction method as described in claim 1, characterized in that, The formula for obtaining the target frame distance matrix is: ； In the formula, The target frame distance matrix, The optimized target frame radar slice image, The distance corresponds to the radar slice image in the i-th frame.

6. The range-gated lidar image reconstruction method as described in claim 1, characterized in that, The formula for obtaining the matrix of the detected target objects is: ； In the formula, The matrix represents the detected target objects, where N is the number of frames in the radar slice image. This is the target frame distance matrix.

7. The range-gated lidar image reconstruction method as described in claim 2, characterized in that, The target frame distance information enhancement map is obtained by performing a bitwise AND operation on the difference image before binarization and the difference image after binarization of the target frame, as shown in the formula: ； In the formula, Enhanced map of target frame distance information. The difference image of the target frame before binarization. The difference image after binarization of the target frame.

8. The range-gated lidar image reconstruction method as described in claim 1, characterized in that, The preprocessing refers to denoising the radar slice images at several different distances to obtain denoised radar slice images.

9. The range-gated lidar image reconstruction method as described in claim 8, characterized in that, The denoising is achieved through median filtering, mean filtering, or Gaussian filtering.

10. The range-gated lidar image reconstruction method as described in claim 1, characterized in that, The pseudo-color processing is JET color mapping, piecewise linear mapping, or rainbow mapping.

Images