Rail transit environment sensing method and rail transit environment sensing device

Abstract

The application discloses an environment sensing method and device for rail transit, and the method comprises the following steps: acquiring continuous frame image information F n ; performing Gaussian filtering processing on the continuous frame image information F n to obtain initial processing image information; performing fusion operation on the initial processing image information to obtain fused image information; performing reprocessing operation on the fused image information to obtain reprocessing image information; and determining environment sensing information of rail transit based on the reprocessing image information. Through data optimization processing and detail enhancement processing on the continuous frame image collected in the rail transit process, the data analysis and processing amount in the environment sensing process is effectively reduced, the environment sensing efficiency is improved, and the accuracy of environment sensing is effectively improved.

Description

Technical Field

[0001] This invention relates to the field of information detection technology, and more specifically to an environmental sensing method and an environmental sensing device for rail transit. Background Technology

[0002] Environmental perception of rail transit lines is a crucial technical means to maintain the safe operation of rail vehicles. There are two existing methods for rail transit line perception: one is through trackside equipment such as infrared laser scanning and trackside cameras; the other is based on onboard equipment such as lidar and cameras.

[0003] Environmental perception methods based on trackside cameras are widely used in the rail transit field because they can directly improve existing rail transit road monitoring systems. Existing environmental perception methods based on trackside cameras generate perception information about the rail transit route by processing and analyzing each frame of the image.

[0004] However, in practical applications, due to the high operating speed of rail transit, and with the continuous development of technology, the speed of rail transit is constantly increasing. High-speed operation also requires a continuous increase in the video acquisition frequency. On the one hand, if every image is processed, there will be a large amount of redundant information, which will greatly increase the amount of computation, waste computing power, and greatly reduce the computational efficiency. On the other hand, if the video acquisition frequency is reduced, high-speed movement will cause the loss of key information, resulting in false detection or missed detection, which cannot meet the actual needs. Summary of the Invention

[0005] To overcome the aforementioned technical problems in the prior art, this invention provides an environmental perception method for rail transit. By performing data optimization and detail enhancement processing on continuous frame images collected during rail transit, the method effectively reduces the amount of data analysis and processing in the environmental perception process, improves environmental perception efficiency, and simultaneously enhances the accuracy of environmental perception.

[0006] To achieve the above objectives, embodiments of the present invention provide an environmental perception method for rail transit, the method comprising: acquiring continuous frame image information; performing Gaussian filtering on the continuous frame image information to obtain pre-processed image information; performing a fusion operation on the pre-processed image information to obtain fused image information; performing a reprocessing operation on the fused image information to obtain reprocessed image information; and determining environmental perception information for rail transit based on the reprocessed image information.

[0007] Preferably, the method further includes: after acquiring the continuous frame image information, acquiring a preset sampling rule; randomly sampling the images in the continuous frame image information based on the preset sampling rule to obtain a sampled image; and using the sampled image as new continuous frame image information.

[0008] Preferably, performing Gaussian filtering on the consecutive frame image information to obtain the initially processed image information includes: determining the consecutive frame image information F n Gaussian parameter information P for each image t The Gaussian parameter information P t Characterized as: P t ={k,σ|k=2t-1,σ=t-0.5}, where, F n ={f t ,f t-1 ,…,f t-n},t∈T + f t The current frame is represented by t, the ordinal number of each image is represented by n, and F is represented by n. n The number of images in the image; based on the Gaussian parameter information P t Determine the corresponding Gaussian distribution matrix G(x,y), which is characterized as... Based on the Gaussian distribution matrix G(x,y) and the continuous frame image information F n Generate the pre-processed image information G n The image information G after initial processing n Characterized as

[0009] Preferably, performing a fusion operation on the initially processed image information to obtain fused image information includes: determining the weighting coefficient w for each image in the initially processed image information. t The weighting coefficient w t Characterized as Based on the weighting coefficient w t Perform a weighted fusion operation on the current frame to obtain the weighted fused image d. t The weighted fused image d t Characterized as d t =∑g t *w t Based on the weighted fused image d t Generate fused image information.

[0010] Preferably, the step of performing a reprocessing operation on the fused image information to obtain reprocessed image information includes: obtaining the standard deviation information Ms of each image in the fused image information; and processing the standard deviation information Ms...s performing a normalization operation to obtain normalized information M norm ; performing a weighting operation on the fused image d t to obtain a reprocessed image d' t , wherein the reprocessed image d' t is represented as d' t = d t · M norm + f t · (1-M norm ); performing a median filtering operation on the reprocessed image d' t to obtain a filtered image; and generating reprocessed image information based on the filtered image.

[0011] Preferably, the obtaining of the standard deviation information M s for each image in the fused image information comprises: performing a color transformation operation on each image in the fused image information to obtain a corresponding transformed image; obtaining luminance data of the transformed image; and determining the standard deviation information M s for each image based on the luminance data and a time sequence T of each image in the fused image information.

[0012] Correspondingly, the present application also provides an environment perception device for rail transit, which comprises: an image acquisition unit configured to acquire continuous frame image information; a first processing unit configured to perform a Gaussian filtering operation on the continuous frame image information to obtain preliminary processed image information; an image fusion unit configured to perform a fusion operation on the preliminary processed image information to obtain fused image information; a second processing unit configured to perform a reprocessing operation on the fused image information to obtain reprocessed image information; and an environment perception unit configured to determine environment perception information for rail transit based on the reprocessed image information.

[0013] Preferably, the device further comprises a sampling unit configured to: after the acquisition of the continuous frame image information, acquire a preset sampling rule; perform random sampling on images in the continuous frame image information based on the preset sampling rule to obtain sampled images; and take the sampled images as new continuous frame image information.

[0014] Preferably, the first processing unit comprises: a parameter information determination module configured to determine Gaussian parameter information P n for each image in the continuous frame image information F t , wherein the Gaussian parameter information P t is represented as: P t = {k, σ | k = 2t-1, σ = t-0.5}, wherein F n = {f tf t-1 , ..., f t-n}, t∈T + f t The current frame is represented by t, the ordinal number of each image is represented by n, and F is represented by n. n The number of images in the data; the Gaussian information determination module, used to determine the Gaussian parameter information P. t Determine the corresponding Gaussian distribution matrix G(x, y), which is characterized as... The Gaussian processing module is used to process the Gaussian distribution matrix G(x, y) and the continuous frame image information F. n Generate the pre-processed image information G n The image information G after initial processing n Characterized as

[0015] Preferably, the image fusion unit includes: a weighting coefficient determination module, used to determine the weighting coefficient w for each image in the pre-processed image information. t The weighting coefficient w t Characterized as The weighted fusion module is used to perform fusion based on the weighting coefficients w. t Perform a weighted fusion operation on the current frame to obtain the weighted fused image d. t The weighted fused image d t Characterized as d t =∑g t *w t The image fusion determination module is used to determine the image based on the weighted fused image d. t Generate fused image information.

[0016] Preferably, the second processing unit includes: a standard deviation acquisition module, used to acquire the standard deviation information M of each image in the fused image information. s The normalization module is used to normalize the standard deviation information M. s Perform a normalization operation to obtain the normalized information M. norm The weighted fusion module is used to process the weighted fused image d according to preset rules. t Processing is performed to obtain the reprocessed image d′ t The reprocessed image d′ t Characterized as d′ t =d t ·M norm +f t ·(1-M norm The median filtering module is used to process the reprocessed image d′. tPerform median filtering processing to obtain a filtered image; a reprocessing image determination module configured to generate reprocessing image information based on the filtered image.

[0017] Preferably, the standard deviation acquisition module is configured to perform a color transformation operation on each of the fused images to obtain a corresponding transformed image, acquire brightness data of the transformed image, and determine standard deviation information M of each of the fused images based on the brightness data and a time sequence T of each of the fused images. s .

[0018] In another aspect, the present application also provides a computer readable storage medium having a computer program stored thereon, the program being executed by a processor to implement the method provided by the present application.

[0019] Through the technical solution provided by the present application, the present application has at least the following technical effects:

[0020] By using the image acquisition device to acquire continuous frame image information in the rail transit transportation process, and by using the Gaussian filtering operation to perform local ignoring processing on the continuous frame image information, the amount of data required to be processed in the environment perception process is effectively reduced, the data processing efficiency is improved, and the real-time and effectiveness of the environment perception are improved.

[0021] Second, by further performing image weighted fusion on the processed image, the information of the current frame is not excessively lost while the local content is ignored, so that the motion information in the continuous frame image is saved in a blurred manner, and the time dimension information of the continuous frame is compressed into the spatial information of a frame image.

[0022] Third, by reprocessing the fused image, the enhancement effect of static details is realized, the pixel dynamic change of the high-speed moving object in the continuous frame image is effectively preserved, the details of the static object without change are preserved, the global motion blur and pixel jitter are reduced, and the accuracy of the environment perception is effectively improved.

[0023] Other features and advantages of the present application will be described in detail in the following specific implementation part. BRIEF DESCRIPTION OF DRAWINGS

[0024] The accompanying drawings are included to provide a further understanding of the present application and constitute a part of the specification, and are used together with the following specific implementation to explain the present application, but do not constitute a limitation on the present application. In the drawings:

[0025] Figure 1 is a specific implementation flowchart of the rail transit environment perception method provided by the present application;

[0026] Figure 2 is a schematic diagram of Gaussian filtering processing of an image in the environment perception method of rail transit provided by the embodiment of the present application;

[0027] Figure 3 is a schematic diagram of calculating standard deviation information of an image in the environment perception method of rail transit provided by the embodiment of the present application;

[0028] Figure 4 is a schematic diagram of reprocessing a fused image in the environment perception method of rail transit provided by the embodiment of the present application;

[0029] Figure 5 is a structural schematic diagram of the environment perception device of rail transit provided by the embodiment of the present application. DETAILED DESCRIPTION

[0030] The specific embodiments of the embodiments of the present application will be described in detail below with reference to the accompanying drawings. It should be understood that the specific embodiments described herein are only used to illustrate and explain the embodiments of the present application, and are not used to limit the embodiments of the present application.

[0031] The terms "system" and "network" in the embodiments of the present application can be used interchangeably. "Multiple" means two or more, and therefore "multiple" in the embodiments of the present application can also be understood as "at least two". "And / or", which describes the association relationship of the associated objects, means that there can be three relationships, for example, A and / or B can mean that there are three cases of A alone, A and B together, and B alone. In addition, the character " / ", if not specially stated, generally represents that the associated objects before and after are in an "or" relationship. In addition, it should be understood that in the description of the embodiments of the present application, "first", "second", etc. are only used for the purpose of distinguishing the description, and cannot be understood as indicating or implying relative importance, nor can it be understood as indicating or implying order.

[0032] Please refer to Figure 1 The embodiment of the present application provides an environment perception method of rail transit, which comprises:

[0033] S10) acquiring continuous frame image information;

[0034] S20) performing Gaussian filtering processing on the continuous frame image information to obtain initial processing image information;

[0035] S30) performing fusion operation on the initial processing image information to obtain fused image information;

[0036] S40) performing reprocessing operation on the fused image information to obtain reprocessing image information;

[0037] S50) determining the environment perception information of the rail transit based on the reprocessed image information.

[0038] In a possible implementation, the continuous frame image information is first acquired, for example, by an image acquisition device arranged beside the rail, and after the rail vehicle moves to the vicinity of the image acquisition device, continuous video information is started to be acquired, the video information is composed of continuous multiple frames of images, and the continuous frame image information F is generated by the continuous frames of images in the acquisition time period T n , for example, the continuous frame image information F including n frames of images n is represented as F n ={f t , f t-1 , …, f t-n}, t∈T + , where t is the current frame number, and f t is the current frame. At this time, in order to reduce the amount of data to be processed, the continuous frame image information F n is subjected to Gaussian filtering processing, and the preprocessed image information is obtained, then the above preprocessed image information is subjected to fusion operation, and the fused image information is obtained, at this time, the fused image information is further subjected to reprocessing operation, for example, the static details of each image in the fused image information are enhanced by reprocessing the fused image information, to obtain the reprocessed image information, at this time, the environment of the rail vehicle is analyzed according to the above reprocessed image information, so as to obtain the corresponding accurate environment perception information.

[0039] In the embodiment of the present application, in the process of perceiving the environment of the rail vehicle, the continuous frame image information acquired is processed by Gaussian filtering operation, so as to greatly reduce the local details in the continuous frame image information, thereby greatly reducing the operation amount and improving the operation efficiency, and at the same time, the reprocessing operation of static detail enhancement is performed on the fused image information, so as to realize further optimization extraction and analysis of the static details, improve the accuracy of perceiving the environment of the rail transit, and improve the utilization rate of operation resources.

[0040] In order to further improve the fusion processing efficiency of the high frame rate video and reduce the operation amount in the fusion processing process, in the embodiment of the present application, the method further includes: after the continuous frame image information is acquired, a preset sampling rule is acquired; the images in the continuous frame image information are randomly sampled based on the preset sampling rule, to obtain the sampled images; and the sampled images are taken as new continuous frame image information.

[0041] In a possible implementation, after the continuous frame image information is acquired, a preset sampling rule is further acquired, for example, the preset sampling rule can be a sampling rule of sampling according to a preset interval, and according to the preset sampling rule, a corresponding frame image is sequentially extracted from the continuous frame image information according to the preset interval as a new continuous frame image, so that more sparse continuous frame images are obtained, and the sampled image is taken as new continuous frame image information.

[0042] In the embodiment of the application, by sampling the originally acquired continuous frame image information, the number of images in the continuous frame image information can be further reduced, so that the data processing amount is further reduced and the processing efficiency is improved on the basis of further improving the fusion processing efficiency of the high frame rate video.

[0043] In the embodiment of the application, the Gaussian filtering processing is performed on the continuous frame image information to obtain the initial processing image information, including: determining Gaussian parameter information P n of each image in the continuous frame image information F t , the Gaussian parameter information P t is represented as: P t ={k, σ | k = 2t-1, σ = t-0.5}, where F n ={f t , f t-1 , …, f t-n}, t ∈ T + , f t represents a current frame, t represents the ordinal number of each image, and n represents the number of images in F n ; determining a corresponding Gaussian distribution matrix G(x, y) based on the Gaussian parameter information P t , the Gaussian distribution matrix G(x, y) is represented as generating the initial processing image information G n based on the Gaussian distribution matrix G(x, y) and the continuous frame image information F n , the initial processing image information G n is represented as

[0044] In a possible implementation, Gaussian filtering processing is performed on each image in the acquired continuous frame image information F n , first, the Gaussian parameter information P t of each image is calculated according to the ordinal number t of the image, for example, the Gaussian parameter information P t is a parameter of a two-dimensional Gaussian distribution matrix of the image, for example, P t ={k, σ | k = 2t-1, σ = t-0.5}, and then the Gaussian parameter information P tFurther determine a Gaussian distribution matrix (Gaussian kernel) G(x, y), for example At this time, each image is subjected to a convolution operation with the corresponding Gaussian distribution matrix, thereby obtaining the processed image information G n Corresponding to the processed image information G n , for example The processed image information G n is a frame sequence with a time forgetting attribute, and the farther away from the current time t, the greater the Gaussian distribution matrix is subjected to a convolution operation, thereby obtaining a "forgotten image" with more blurred local details. Please refer to Figure 2 , the processed image information G n of an image provided by an embodiment of the present application after Gaussian filtering, according to the drawing, it can be seen that after Gaussian filtering, the local details in the image become more blurred, thereby effectively reducing the amount of data to be processed in the image and improving the operation efficiency.

[0045] In actual application, if each image in the processed image information G n is subjected to separate analysis and processing, the time correlation of the spatial characteristics of different images cannot be reflected, therefore, in order to express stronger time correlation in the process of fusing the continuous frame images, a further weighted image fusion operation is adopted to fuse the continuous frame images.

[0046] In an embodiment of the present application, the fusion operation on the processed image information is performed to obtain the fused image information, comprising: determining a weighting coefficient w t of each image in the processed image information, wherein the weighting coefficient w t is represented as Based on the weighting coefficient w t , a weighted fusion operation is performed on the current frame to obtain a weighted fused image d t , wherein the weighted fused image d t is represented as d t =∑g t *w t ; and the fused image information is generated based on the weighted fused image d t .

[0047] In a possible implementation, first, a weighting coefficient w t of each image in the processed image information is obtained, for example Then, each image is subjected to a weighted processing based on the weighting coefficient w t , thereby obtaining a weighted fused image d t , for example d t =∑g t *w tat this time, the weighted fused image d t corresponding fused image information is generated, for example, the fused image information is a set of all weighted fused images d t .

[0048] In the embodiment of the present application, by using the weighted fusion manner to further fuse each image, the change information of each pixel in the previous frame relative to the current frame is included in each image, that is, the pixel dynamic change in the limited sequence is compressed into one frame of image, thereby greatly improving the time correlation of each image in the spatial characteristics, and improving the perception accuracy in the subsequent environment perception process.

[0049] In the actual application process, although the image in the above fused image information contains a certain degree of pixel dynamic change information, the above information is global information, and thus dynamic blur of the image is caused, thereby causing certain disturbance to the subsequent environment perception analysis and recognition.

[0050] To solve the above technical problem, in the embodiment of the present application, the reprocessing operation is performed on the fused image information to obtain reprocessed image information, including: acquiring standard deviation information M s of each image in the fused image information; performing normalization operation on the standard deviation information M s to obtain normalized information M norm ; processing the weighted fused image d t according to a preset rule to obtain a reprocessed image d' t , the reprocessed image d' t is represented as d' t =d t ·M norm +f t ·(1-M norm ); performing median filtering processing on the reprocessed image d' t to obtain a filtered image; and generating reprocessed image information based on the filtered image.

[0051] Further, in the embodiment of the present application, the standard deviation information M s of each image in the fused image information is acquired, including: performing color transformation operation on each image in the fused image to obtain a corresponding transformed image; acquiring brightness data of the transformed image; and determining the standard deviation information M s of each image based on the brightness data and the time sequence T of each image in the fused image.

[0052] In one possible implementation, before performing environmental perception, the standard deviation information M of each image in the fused image information is further obtained. s Specifically, firstly, a color transformation operation is performed on each image in the fused image to obtain the corresponding transformed image. For example, the RGB colors of each image are first transformed to the HSV space, which is a model that includes hue, saturation, and value color parameters. Then, the value data is further extracted, and the standard deviation information M of each image is determined based on the value data and the time series T of each image. s For example, the brightness data mentioned above can be used as the data for calculating the standard deviation mask, and the standard deviation of the pixel value at each position along the time sequence T in each image can be calculated in the T-dimensional direction, thereby obtaining the standard deviation information M of each image. s Please see Figure 3 The standard deviation information M for each image is provided in the embodiments of the present invention. s A schematic diagram.

[0053] In this embodiment of the invention, the standard deviation of each pixel position in each image is determined by calculating the standard deviation of each image, that is, the degree of brightness change of each pixel along the temporal direction in the fused image information is determined, for example, in the attached... Figure 3 As can be seen, after standard deviation processing, static objects show smaller grayscale changes. For example, roads, skies, and buildings in the image have large areas of the same grayscale value, and their grayscale changes between frames are very small compared to moving objects, such as cars.

[0054] To further reduce global pixel jitter and blurring, the image after calculating the standard deviation is further processed. For example, the standard deviation information M is... s Perform a normalization operation to obtain the normalized information M. norm Then, the weighted fused image d is processed according to preset rules. t Processing is performed to obtain the reprocessed image d′ t For example, d′ t =d t ·M norm +f t ·(1-M norm In this embodiment of the invention, the weighted fused image d is calculated using the above-described calculation rules. t With the corresponding original current frame image f t The image f in the current frame was fused. t The details covered in the weighted fused image d tThis effectively reduces global pixel jitter and blurring in areas with minimal dynamic changes, thus enhancing static details. Please refer to [link to relevant documentation]. Figure 4 This is an embodiment of the present invention that provides a reprocessed image d′ obtained by reprocessing the fused image. t A schematic diagram.

[0055] At this point, the reprocessed image d′ is further processed. t Perform median filtering to remove the median filter from the reprocessed image d′. t To further improve the accuracy and reliability of each image data by removing particulate noise, and to obtain filtered images, reprocessed image information is generated based on each filtered image, which is a set of filtered images. At this time, environmental perception operations for rail transit are performed based on the above reprocessed image information, so as to quickly and accurately determine the environmental perception information of rail transit.

[0056] In this embodiment of the invention, by performing data optimization and detail enhancement processing on multiple consecutive images collected during the operation of rail transit, the amount of data that needs to be processed during environmental perception can be effectively reduced, the efficiency of environmental perception can be improved, and the accuracy of dynamic detail perception can be effectively improved, thereby enhancing the operational safety of rail transit.

[0057] The environmental sensing device for rail transit provided in the embodiments of the present invention will be described below with reference to the accompanying drawings.

[0058] Please see Figure 5 Based on the same inventive concept, this invention provides an environmental sensing device for rail transit. The device includes: an image acquisition unit for acquiring continuous frame image information; a first processing unit for performing Gaussian filtering on the continuous frame image information to obtain pre-processed image information; an image fusion unit for performing a fusion operation on the pre-processed image information to obtain fused image information; a second processing unit for performing a reprocessing operation on the fused image information to obtain reprocessed image information; and an environmental sensing unit for determining environmental sensing information of rail transit based on the reprocessed image information.

[0059] In this embodiment of the invention, the device further includes a sampling unit, which is used to: after acquiring continuous frame image information, acquire a preset sampling rule; randomly sample the images in the continuous frame image information based on the preset sampling rule to obtain a sampled image; and use the sampled image as new continuous frame image information.

[0060] In this embodiment of the invention, the first processing unit includes: a parameter information determination module, configured to determine the continuous frame image information F. nGaussian parameter information P of each image t , the Gaussian parameter information P t is characterized as: P t ={k, sigma | k=2t-1, sigma=t-0.5}, wherein F n ={f t , f t-1 , …, f t-n}, t∈T + , f t characterizes a current frame, t characterizes the ordinal number of each image, and n characterizes the number of images in F n ; a Gaussian information determination module is configured to determine a corresponding Gaussian distribution matrix G(x, y) based on the Gaussian parameter information P t , and the Gaussian distribution matrix G(x, y) is characterized as A Gaussian processing module is configured to generate the initial processed image information G n based on the Gaussian distribution matrix G(x, y) and the continuous frame image information F n , and the initial processed image information G n is characterized as

[0061] In the embodiment of the present application, the image fusion unit comprises: a weighting coefficient determination module configured to determine a weighting coefficient w t of each image in the initial processed image information, and the weighting coefficient w t is characterized as A weighted fusion module is configured to perform a weighted fusion operation on the current frame based on the weighting coefficient w t to obtain a weighted fused image d t , and the weighted fused image d t is characterized as d t =∑g t *w t ; a fused image determination module is configured to generate fused image information based on the weighted fused image d t .

[0062] In the embodiment of the present application, the second processing unit comprises: a standard deviation acquisition module configured to acquire standard deviation information M s of each image in the fused image information; a normalization module configured to perform a normalization operation on the standard deviation information M s to obtain normalized information M norm ; and a weighted fusion module configured to process the weighted fused image d t according to a preset rule to obtain a reprocessed image d′ t , and the reprocessed image d′ t is characterized as d′t = d t · M norm + f t · (1-M norm ); a median filter module, configured to perform median filtering processing on the reprocessed image d' t to obtain a filtered image; and a reprocessed image determination module, configured to generate reprocessed image information based on the filtered image.

[0063] In the embodiment of the present application, the standard deviation acquisition module is configured to: perform color transformation on each of the fused images to obtain a corresponding transformed image; acquire brightness data of the transformed image; and determine standard deviation information M s of each of the fused images based on the brightness data and a time sequence T of each of the fused images.

[0064] Further, the embodiment of the present application also provides a computer readable storage medium, which stores a computer program, and the program is executed by a processor to implement the method in the embodiment of the present application.

[0065] The optional implementation manners of the embodiment of the present application are described in detail above in combination with the drawings, however, the embodiment of the present application is not limited to the specific details in the above implementation manners, and various simple modifications can be made to the technical solution of the embodiment of the present application within the technical concept of the embodiment of the present application, and these simple modifications all belong to the protection scope of the embodiment of the present application.

[0066] In addition, it should be noted that each specific technical feature described in the above specific implementation manners can be combined in any appropriate manner without contradiction. In order to avoid unnecessary repetition, the embodiment of the present application will not further describe various possible combination manners.

[0067] Those skilled in the art can understand that all or part of the steps of the methods in the above embodiments can be completed by a program instructing related hardware, the program is stored in a storage medium, and includes a plurality of instructions for causing a single-chip microcomputer, a chip or a processor to execute all or part of the steps of the methods described in various embodiments of the present application. The foregoing storage medium includes a U disk, a mobile hard disk, a read-only memory (ROM), a random access memory (RAM), a magnetic disk or an optical disk and various program code storage media.

[0068] In addition, various different implementation manners of the embodiment of the present application can also be combined in any appropriate manner, as long as it does not contradict the technical concept of the embodiment of the present application, it should also be considered as the disclosed content of the embodiment of the present application.

Claims

1. A method for environmental perception of rail transit, characterized in that, The method comprises: obtaining continuous frame image information; performing Gaussian filtering processing on the continuous frame image information to obtain image information after initial processing; performing fusion operation on the image information after initial processing to obtain image information after fusion; performing reprocessing operation on the image information after fusion to obtain reprocessed image information; determining environment perception information of rail transit based on the reprocessed image information; wherein the performing reprocessing operation on the image information after fusion to obtain reprocessed image information comprises: obtaining standard deviation information of each image in the fused image information ; to the standard deviation information performing a normalization operation to obtain normalized information ; performing processing on the weighted fused image according to a preset rule to obtain a reprocessed image performing processing on the weighted fused image according to a preset rule to obtain a reprocessed image , the reprocessed image is characterized by , is characterized by a current frame; reprocessing the image performing a median filter process to obtain a filtered image generating reprocessed image information based on the filtered image.

2. The method of claim 1, wherein, The method further comprises: after the obtaining continuous frame image information, obtaining a preset sampling rule; based on the preset sampling rule, randomly sampling images in the continuous frame image information to obtain sampled images; taking the sampled images as new continuous frame image information.

3. The method of claim 1, wherein, The performing Gaussian filtering processing on the continuous frame image information to obtain image information after initial processing comprises: determining the continuous frame image information gaussian parameter information of each image gaussian parameter information characterized by gaussian parameter information wherein t represents the ordinal number of each image, n represents the number of images in represents a set of positive integers;​ based on the gaussian parameter information determining a corresponding gaussian distribution matrix , the gaussian distribution matrix characterized in that ; based on the gaussian distribution matrix with the continuous frame image information generating the preliminary processed image information , the preliminary processed image information characterized by .

4. The method of claim 3, wherein, The performing fusion operation on the image information after initial processing to obtain image information after fusion comprises: determining a weighting factor for each of the images in the pre-processed image information , the weighting factor is characterized in that ; based on the weighting coefficients performing a weighted fusion operation on the current frame to obtain a weighted fused image , the weighted fused image characterized by ; based on the weighted fused image generating fused image information.

5. The method of claim 1, wherein, The standard deviation information of each image in the fused image information is acquired comprising: performing color transformation operation on each image in the image information after fusion to obtain corresponding transformed images; obtaining brightness data of the transformed images. determining standard deviation information of each of the images based on the luminance data and a time sequence T of each of the images after the fusion .

6. An environmental perception device for rail transit, characterized in that, The device comprises: an image obtaining unit configured to obtain continuous frame image information; a first processing unit configured to perform Gaussian filtering processing on the continuous frame image information to obtain image information after initial processing; an image fusion unit configured to perform fusion operation on the image information after initial processing to obtain image information after fusion; a second processing unit configured to perform reprocessing operation on the image information after fusion to obtain reprocessed image information; an environment perception unit configured to determine environment perception information of rail transit based on the reprocessed image information; wherein the second processing unit comprises: The standard deviation acquisition module is used to acquire the standard deviation information of each image in the fused image information. The normalization module is used to normalize the standard deviation information. Perform a normalization operation to obtain normalized information. The weighted fusion module is used to weight and fuse the image according to preset rules. Process the image to obtain a reprocessed image. The reprocessed image Characterized as , The current frame is represented by a median filtering module, used for reprocessing the image. A median filtering process is performed to obtain a filtered image; a reprocessed image determination module is used to generate reprocessed image information based on the filtered image.

7. The apparatus of claim 6, wherein, The device further comprises a sampling unit configured to: after the obtaining continuous frame image information, obtain a preset sampling rule; based on the preset sampling rule, randomly sample images in the continuous frame image information to obtain sampled images; take the sampled images as new continuous frame image information.

8. The apparatus of claim 6, wherein, The first processing unit comprises: A parameter information determining module is configured to determine the continuous frame image information Gaussian parameter information of each image The Gaussian parameter information is characterized by: , wherein , t represents the ordinal number of each image, n represents the number of images in the set represents a set of positive integers; a Gaussian information determination module, configured to determine Gaussian parameter information based on the Gaussian distribution matrix determine a corresponding Gaussian distribution matrix , the Gaussian distribution matrix characterized by ; a Gaussian processing module configured to generate a Gaussian distribution matrix based on the plurality of image information with the continuous frame image information generate the preliminary processed image information , the preliminary processed image information characterized by .

9. The apparatus of claim 8, wherein, The image fusion unit comprises: A weighting coefficient determination module is configured to determine a weighting coefficient of each image in the pre-processed image information The weighting coefficient is characterized by ; The weighted fusion module is configured to fuse the first image and the second image based on the weighting coefficients perform a weighted fusion operation on the current frame to obtain a weighted fusion image , the weighted fusion image characterized by ; a fusion image determination module configured to determine a fusion image based on the weighted fused image generate fused image information.

10. The apparatus of claim 6, wherein, The standard deviation obtaining module is configured to: perform color transformation operation on each image in the image information after fusion to obtain corresponding transformed images; obtain brightness data of the transformed images. determining standard deviation information of each of the images based on the luminance data and a time sequence T of each of the images after the fusion .

11. A computer readable storage medium having stored thereon a computer program, characterized in that The program is executed by the processor to implement the method in any one of claims 1-5.

Images