Vehicle-mounted multi-camera joint calibration method and device

Abstract

The invention provides a vehicle-mounted multi-camera joint calibration method and device, and relates to the technical field of vehicles. When the method is executed, firstly, multi-camera input images corresponding to a plurality of frames respectively are obtained, then feature extraction is carried out on the multi-camera input images corresponding to the plurality of frames respectively to obtain feature sets corresponding to the plurality of frames respectively, then, according to the feature sets corresponding to the plurality of frames respectively, a plurality of traveling distances of a vehicle between two frames are determined, and finally, the traveling distances of the vehicle between the two frames are determined. And calculating a camera pose by using the advancing distances of the plurality of vehicles between the two frames, so as to determine a camera calibration result according to the camera pose. In this way, through the feature sets corresponding to the multiple frames, the multiple advancing distances of the vehicle between the two frames are determined, and then the camera pose is obtained through calculation, so that the camera calibration result can be accurately determined through the camera pose. Therefore, the vehicle-mounted multi-camera joint calibration effect can be improved.

Description

Technical Field

[0001] The present application relates to the field of vehicle technology, and in particular to a vehicle-mounted multi-camera joint calibration method and device. Background Art

[0002] Multi-camera calibration technology is usually used when the whole vehicle is off the production line. The vehicle will be driven to a designated calibration room, and multi-camera calibration will be performed through the already arranged calibration room and the calibration blocks with fixed patterns. Only the calibrated cameras can present seamless panoramic surround images, and use this calibration result to provide the automatic parking function with parking space, obstacles and other distance information from the vehicle, thereby supporting the automatic driving function. If there is no calibration or the calibration is inaccurate, the image will be broken or deformed, resulting in poor imaging effect and low ranging accuracy. In the prior art, the joint calibration of multiple cameras has a poor effect.

[0003] In summary, how to improve the joint calibration effect of vehicle-mounted multi-cameras is a problem that technical personnel in this field need to solve urgently. Summary of the invention

[0004] In view of this, the present application provides a vehicle-mounted multi-camera joint calibration method and device, aiming to improve the vehicle-mounted multi-camera joint calibration effect.

[0005] In a first aspect, the present application provides a vehicle-mounted multi-camera joint calibration method, comprising:

[0006] Obtain input images of multiple cameras corresponding to multiple frames;

[0007] Extracting features from input images of multiple cameras corresponding to the multiple frames to obtain feature sets corresponding to the multiple frames;

[0008] Determining a plurality of travel distances of vehicles between two frames according to the feature sets respectively corresponding to the plurality of frames;

[0009] Calculating the camera pose using the travel distances of the multiple vehicles between two frames;

[0010] A camera calibration result is determined according to the camera pose.

[0011] Optionally, extracting features from the input images of multiple cameras corresponding to the multiple frames to obtain feature sets corresponding to the multiple frames respectively includes:

[0012] The Transformer model is used to extract features from input images of multiple cameras corresponding to the multiple frames, so as to obtain feature sets corresponding to the multiple frames.

[0013] Optionally, determining a plurality of vehicle travel distances between two frames according to the feature sets respectively corresponding to the plurality of frames includes:

[0014] Match the feature sets corresponding to the two frames using the optical flow method to obtain the travel distance of the vehicle between the two frames;

[0015] Repeat the step of matching the feature sets corresponding to the two frames respectively using the optical flow method to obtain the travel distance of the vehicle between the two frames, and obtain multiple travel distances of the vehicle between the two frames.

[0016] Optionally, performing optical flow matching on the feature sets corresponding to the two frames to obtain the travel distance of the vehicle between the two frames includes:

[0017] The feature sets corresponding to two consecutive frames are matched using the optical flow method to obtain the travel distance of the vehicle between the two consecutive frames.

[0018] Optionally, after extracting features from the input images of the multiple cameras corresponding to the multiple frames to obtain feature sets corresponding to the multiple frames, the method further includes:

[0019] According to the feature set corresponding to each frame, the camera features in each frame are matched in pairs to obtain the matching result corresponding to each frame.

[0020] Optionally, before determining the camera calibration result according to the camera pose, the method further includes:

[0021] Using the matching results corresponding to each frame, adjusting the camera pose to obtain an adjusted camera pose;

[0022] The determining of the camera calibration result according to the camera posture comprises:

[0023] A camera calibration result is determined according to the adjusted camera pose.

[0024] Optionally, performing optical flow matching on the feature sets corresponding to the two frames to obtain the travel distance of the vehicle between the two frames includes:

[0025] Acquire vehicle wheel speed device signals corresponding to the two frames respectively;

[0026] Performing optical flow matching on the feature sets corresponding to the two frames respectively, to obtain optical flow matching results corresponding to the two frames;

[0027] The travel distance of the vehicle between the two frames is determined according to the vehicle wheel speed signal corresponding to the two frames and the optical flow matching results corresponding to the two frames.

[0028] In a second aspect, the present application provides a vehicle-mounted multi-camera joint calibration device, comprising:

[0029] An acquisition module, used to acquire input images of multiple cameras corresponding to multiple frames;

[0030] A feature extraction module, used to extract features from input images of multiple cameras corresponding to the multiple frames, to obtain feature sets corresponding to the multiple frames;

[0031] A first determination module, configured to determine a plurality of travel distances of vehicles between two frames according to feature sets corresponding to the plurality of frames respectively;

[0032] A calculation module, used for calculating the camera pose using the travel distance of the multiple vehicles between two frames;

[0033] The second determination module is used to determine the camera calibration result according to the camera posture.

[0034] Optionally, the feature extraction module includes:

[0035] The feature extraction unit is used to use the Transformer model to extract features from the input images of the multiple cameras corresponding to the multiple frames, so as to obtain feature sets corresponding to the multiple frames.

[0036] Optionally, the first determining module includes:

[0037] The matching unit is used to perform optical flow matching on the feature sets corresponding to the two frames to obtain the travel distance of the vehicle between the two frames; repeat the step of performing optical flow matching on the feature sets corresponding to the two frames to obtain the travel distance of the vehicle between the two frames, and obtain multiple travel distances of the vehicle between the two frames.

[0038] Optionally, the matching unit includes:

[0039] The first matching subunit is used to perform optical flow matching on the feature sets corresponding to two consecutive frames to obtain the travel distance of the vehicle between the two consecutive frames.

[0040] Optionally, the device further comprises:

[0041] The matching module is used to match the camera features in each frame in pairs according to the feature set corresponding to each frame to obtain the matching result corresponding to each frame.

[0042] Optionally, the device further comprises:

[0043] An adjustment module, used to adjust the camera posture using the matching result corresponding to each frame to obtain an adjusted camera posture;

[0044] The second determining module includes:

[0045] The determination unit is used to determine the camera calibration result according to the adjusted camera posture.

[0046] Optionally, the matching unit includes:

[0047] An acquisition subunit, used for acquiring vehicle wheel speed device signals corresponding to the two frames respectively;

[0048] A second matching subunit is used to perform optical flow matching on the feature sets corresponding to the two frames respectively, to obtain optical flow matching results corresponding to the two frames;

[0049] The determination subunit is used to determine the travel distance of the vehicle between the two frames according to the vehicle wheel speed signal corresponding to the two frames and the optical flow matching results corresponding to the two frames.

[0050] The present application provides a method for joint calibration of multiple cameras on a vehicle. When executing the method, first obtain the input images of multiple cameras corresponding to multiple frames, then extract features of the input images of multiple cameras corresponding to multiple frames, and obtain feature sets corresponding to multiple frames. Then, according to the feature sets corresponding to multiple frames, determine the travel distances of multiple vehicles between two frames. Finally, use the travel distances of multiple vehicles between two frames to calculate the camera pose, so as to determine the camera calibration result according to the camera pose. In this way, the travel distances of multiple vehicles between two frames are determined by the feature sets corresponding to multiple frames, and then the camera pose is calculated, so that the camera calibration result can be accurately determined using the camera pose. In this way, the joint calibration effect of multiple cameras on a vehicle can be improved. BRIEF DESCRIPTION OF THE DRAWINGS

[0051] In order to more clearly illustrate the technical solutions in this embodiment or the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0052] Figure 1 A flowchart of a vehicle-mounted multi-camera joint calibration method provided in an embodiment of the present application;

[0053] Figure 2 A schematic diagram of the structure of a vehicle-mounted multi-camera joint calibration device provided in an embodiment of the present application. DETAILED DESCRIPTION

[0054] The following will be combined with the drawings in the embodiments of the present application to clearly and completely describe the technical solutions in the embodiments of the present application. The present application provides a vehicle-mounted multi-camera joint calibration method and device for use in the field of vehicle technology. The above is only an example and does not limit the application field of the method and device provided in the present application.

[0055] Multi-camera calibration technology is usually used when the whole vehicle is off the production line. The vehicle will be driven to a designated calibration room, and multi-camera calibration will be performed through the already arranged calibration room and the calibration blocks with fixed patterns. Only the calibrated cameras can present seamless panoramic surround images, and use this calibration result to provide the automatic parking function with parking space, obstacles and other distance information from the vehicle, thereby supporting the automatic driving function. If there is no calibration or the calibration is inaccurate, the image will be broken or deformed, resulting in poor imaging effect and low ranging accuracy. In the prior art, the joint calibration of multiple cameras has a poor effect.

[0056] After research, the inventor proposed the technical solution of the present application. First, the input images of multiple cameras corresponding to multiple frames are obtained, and then the input images of multiple cameras corresponding to multiple frames are subjected to feature extraction to obtain feature sets corresponding to multiple frames. Then, the travel distances of multiple vehicles between two frames are determined according to the feature sets corresponding to multiple frames. Finally, the travel distances of multiple vehicles between two frames are calculated using the travel distances of multiple vehicles between two frames, so as to determine the camera calibration results according to the camera poses. In this way, the travel distances of multiple vehicles between two frames are determined by the feature sets corresponding to multiple frames, and then the camera poses are calculated, so that the camera calibration results can be accurately determined using the camera poses. In this way, the joint calibration effect of multiple on-board cameras can be improved.

[0057] In order to enable those skilled in the art to better understand the present application, the present application is further described in detail below in conjunction with the accompanying drawings and specific implementation methods. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in the field without making creative work are within the scope of protection of the present application. It should be noted that, for ease of description, only the parts related to the relevant invention are shown in the accompanying drawings. In the absence of conflict, the embodiments in the present application and the features in the embodiments can be combined with each other.

[0058] See also Figure 1 , Figure 1 A flowchart of a vehicle-mounted multi-camera joint calibration method provided in an embodiment of the present application includes:

[0059] S101: Acquire input images of multiple cameras corresponding to multiple frames.

[0060] For example, in this embodiment, a total of N frames of input images need to be obtained, and the multi-camera input image corresponding to the first frame, the multi-camera input image corresponding to the second frame, ..., the multi-camera input image corresponding to the N-1th frame, and the multi-camera input image corresponding to the Nth frame are obtained, where N is a positive integer. The multi-camera input images generally include a front camera input image, a left camera input image, a rear camera input image, and a right camera input image.

[0061] In some implementations, the multiple frames may be consecutive frames, and N may be 15-20.

[0062] S102: Extract features from input images of multiple cameras corresponding to the multiple frames to obtain feature sets corresponding to the multiple frames.

[0063] The Transformer network is used to extract features from the input images of multiple cameras corresponding to multiple frames, and the feature sets corresponding to multiple frames are obtained. The Transformer network is used to extract the salient features of each camera image, and each salient feature is described by different feature vectors. In this way, the semantic description of the features extracted based on the Transformer network is richer and more robust. Moreover, such features are not limited to lane lines, but can also be ground gaps, crosswalks, grid lines, parking spaces, speed bumps, ground arrows, etc., which are completely natural semantics and do not require artificial design. There will be a lot of such features in the common viewing area of ​​two cameras for matching, so that the image fusion effect will be better. Moreover, the features extracted based on the Transformer network are not just color space and shape features, but multi-dimensional and multi-level features combined with local semantics, which have a certain anti-interference ability to the influence of light, brightness, and water accumulation.

[0064] S103: Determine travel distances of multiple vehicles between two frames according to feature sets corresponding to the multiple frames.

[0065] For example, in order to determine the distance traveled by the vehicle between the first frame and the second frame, the optical flow method is first used to extract the position change of the feature points in the feature set corresponding to the first frame and the second frame respectively in the image, and the physical distance corresponding to this change is the actual distance traveled by the vehicle between the two frames. The actual distance traveled by the vehicle between the two frames can be calculated by finding the timestamps of the first frame and the second frame image capture, and the vehicle wheel speedometer signals corresponding to the first frame and the second frame are calculated to obtain the travel distance value.

[0066] S104: Calculate the camera pose using the travel distances of multiple vehicles between two frames.

[0067] In the manner of step S103, the travel distance of a vehicle between two frames can be determined, and the position and posture of the camera can be calculated through the travel distances of multiple vehicles between two frames.

[0068] S105: Determine a camera calibration result according to the camera position and posture.

[0069] Before executing step S105, the camera features in each frame may be matched in pairs according to the feature set corresponding to each frame to obtain the matching result corresponding to each frame. Then, the camera posture is optimized using the matching result corresponding to each frame, and finally the camera calibration result is determined using the optimized camera posture.

[0070] In an embodiment of the present application, the input images of multiple cameras corresponding to multiple frames are first obtained, and then the input images of multiple cameras corresponding to multiple frames are subjected to feature extraction to obtain feature sets corresponding to multiple frames. Then, based on the feature sets corresponding to multiple frames, the travel distances of multiple vehicles between two frames are determined. Finally, the travel distances of multiple vehicles between two frames are used to calculate the camera pose, so as to determine the camera calibration result according to the camera pose. In this way, the travel distances of multiple vehicles between two frames are determined by the feature sets corresponding to multiple frames, and then the camera pose is calculated, so that the camera calibration result can be accurately determined using the camera pose. In this way, the joint calibration effect of multiple on-board cameras can be improved.

[0071] The above introduces the vehicle-mounted multi-camera joint calibration method provided by the embodiment of the present application. The following is an exemplary description of the vehicle-mounted multi-camera joint calibration method in combination with a specific application scenario.

[0072] First, use a multi-camera to obtain 20 consecutive frames of input images, where the multi-camera includes a front camera, a rear camera, a left camera, and a right camera. That is, obtain the multi-camera input image corresponding to the first frame of the continuous frame, the multi-camera input image corresponding to the second frame of the continuous frame, the multi-camera input image corresponding to the third frame of the continuous frame, ..., the multi-camera input image corresponding to the 19th frame of the continuous frame, and the multi-camera input image corresponding to the 20th frame of the continuous frame. Then, use the Transformer network to extract the multi-camera input image corresponding to each continuous frame to obtain the feature set corresponding to each continuous frame, that is, feature set 1 corresponding to the first frame, feature set 2 corresponding to the second frame, feature set 3 corresponding to the third frame, feature set 4 corresponding to the fourth frame, ..., feature set 19 corresponding to the 19th frame, and feature set 20 corresponding to the 20th frame. Then, use the streamer method to match the feature sets of the two consecutive frames to obtain the matching results, and then determine the travel distance of the vehicle in the two consecutive frames according to the vehicle wheel speed meter signal. For example, the streamer method is used to match the feature sets corresponding to the first frame and the second frame respectively, and the matching results of the feature set of the first frame and the feature set of the second frame are obtained. According to the timestamps of the first frame and the second frame, the vehicle's travel distance between the first frame and the second frame is determined using the vehicle's wheel speed meter signal. In this way, 20 sets of vehicle travel distances can be obtained, and the camera pose can be calculated through these 20 sets of vehicle travel distances. Then, for each consecutive frame, the pairwise camera features in the feature set of each consecutive frame are matched to obtain the feature matching results corresponding to each frame, and the camera pose is optimized using each feature matching result. Finally, the camera calibration result is determined using the optimized camera pose.

[0073] The above are some specific implementations of the vehicle-mounted multi-camera joint calibration method provided by the embodiment of the present application. Based on this, the present application also provides a corresponding device. The device provided by the embodiment of the present application will be introduced from the perspective of functional modularization.

[0074] See also Figure 2 , Figure 2 The present invention provides a schematic diagram of the structure of a vehicle-mounted multi-camera joint calibration device according to an embodiment of the present invention. The vehicle-mounted multi-camera joint calibration device 200 includes:

[0075] An acquisition module 210 is used to acquire input images of multiple cameras corresponding to multiple frames;

[0076] A feature extraction module 220 is used to extract features from the input images of the multiple cameras corresponding to the multiple frames to obtain feature sets corresponding to the multiple frames;

[0077] A first determination module 230, configured to determine a plurality of travel distances of vehicles between two frames according to the feature sets respectively corresponding to the plurality of frames;

[0078] A calculation module 240, configured to calculate a camera pose using the travel distances of the plurality of vehicles between two frames;

[0079] The second determination module 250 is used to determine the camera calibration result according to the camera posture.

[0080] Optionally, the feature extraction module 220 includes:

[0081] The feature extraction unit is used to use the Transformer model to extract features from the input images of the multiple cameras corresponding to the multiple frames, so as to obtain feature sets corresponding to the multiple frames.

[0082] Optionally, the first determining module 230 includes:

[0083] The matching unit is used to perform optical flow matching on the feature sets corresponding to the two frames to obtain the travel distance of the vehicle between the two frames; repeat the step of performing optical flow matching on the feature sets corresponding to the two frames to obtain the travel distance of the vehicle between the two frames, and obtain multiple travel distances of the vehicle between the two frames.

[0084] Optionally, the matching unit includes:

[0085] The first matching subunit is used to perform optical flow matching on the feature sets corresponding to two consecutive frames to obtain the travel distance of the vehicle between the two consecutive frames.

[0086] Optionally, the device 200 further includes:

[0087] The matching module is used to match the camera features in each frame in pairs according to the feature set corresponding to each frame to obtain the matching result corresponding to each frame.

[0088] Optionally, the device 200 further includes:

[0089] An adjustment module, used to adjust the camera posture using the matching result corresponding to each frame to obtain an adjusted camera posture;

[0090] The second determining module 250 includes:

[0091] The determination unit is used to determine the camera calibration result according to the adjusted camera posture.

[0092] Optionally, the matching unit includes:

[0093] An acquisition subunit, used for acquiring vehicle wheel speed device signals corresponding to the two frames respectively;

[0094] A second matching subunit is used to perform optical flow matching on the feature sets corresponding to the two frames respectively, to obtain optical flow matching results corresponding to the two frames;

[0095] The determination subunit is used to determine the travel distance of the vehicle between the two frames according to the vehicle wheel speed signal corresponding to the two frames and the optical flow matching results corresponding to the two frames.

[0096] The embodiments of the present application also provide corresponding devices and computer storage media for implementing the solutions provided by the embodiments of the present application.

[0097] The device includes a memory and a processor, the memory is used to store instructions or codes, and the processor is used to execute the instructions or codes so that the device executes the method described in any embodiment of the present application.

[0098] The computer storage medium stores codes, and when the codes are executed, a device executing the codes implements the method described in any embodiment of the present application.

[0099] Through the description of the above implementation methods, it can be known that those skilled in the art can clearly understand that all or part of the steps in the above-mentioned embodiment method can be implemented by means of software plus a general hardware platform. Based on such an understanding, the technical solution of the present application can be embodied in the form of a software product, which can be stored in a storage medium, such as a read-only memory (ROM) / RAM, a magnetic disk, an optical disk, etc., including a number of instructions for a computer device (which can be a personal computer, a server, or a network communication device such as a router) to execute the methods described in each embodiment of the present application or some parts of the embodiments.

[0100] It should be noted that, in this article, relational terms such as first and second, etc. are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "include", "comprise" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device. In the absence of further restrictions, the elements defined by the sentence "comprise a ..." do not exclude the existence of other identical elements in the process, method, article or device including the elements.

[0101] It should also be noted that the various embodiments in this specification are described in a progressive manner, and the same and similar parts between the various embodiments can refer to each other, and each embodiment focuses on the differences from other embodiments. In particular, for the equipment and device embodiments, since they are basically similar to the method embodiments, the description is relatively simple, and the relevant parts can refer to the partial description of the method embodiment. The device embodiments described above are merely schematic, in which the units described as separate components may or may not be physically separated, and the components prompted as units may or may not be physical units, that is, they may be located in one place, or they may be distributed on multiple network units. Some or all of the modules can be selected according to actual needs to achieve the purpose of the scheme of this embodiment. Ordinary technicians in this field can understand and implement it without paying creative labor.

[0102] The above is only a specific implementation of the present application, but the protection scope of the present application is not limited thereto. Any changes or substitutions that can be easily thought of by a person skilled in the art within the technical scope disclosed in the present application should be included in the protection scope of the present application. Therefore, the protection scope of the present application should be based on the protection scope of the claims.

Claims

1. A vehicle-mounted multi-camera joint calibration method, characterized in that: include: Obtain input images of multiple cameras corresponding to multiple frames; Extracting features from input images of multiple cameras corresponding to the multiple frames to obtain feature sets corresponding to the multiple frames; Determining a plurality of travel distances of vehicles between two frames according to the feature sets respectively corresponding to the plurality of frames; Calculating the camera pose using the travel distances of the multiple vehicles between two frames; A camera calibration result is determined according to the camera pose.

2. The method according to claim 1, characterized in that The step of extracting features from the input images of the multiple cameras corresponding to the multiple frames to obtain feature sets corresponding to the multiple frames includes: The Transformer model is used to extract features from input images of multiple cameras corresponding to the multiple frames, so as to obtain feature sets corresponding to the multiple frames.

3. The method according to claim 1, characterized in that: The step of determining a plurality of vehicle travel distances between two frames according to the feature sets respectively corresponding to the plurality of frames comprises: Match the feature sets corresponding to the two frames using the optical flow method to obtain the travel distance of the vehicle between the two frames; Repeat the step of matching the feature sets corresponding to the two frames respectively using the optical flow method to obtain the travel distance of the vehicle between the two frames, and obtain multiple travel distances of the vehicle between the two frames.

4. The method according to claim 3, characterized in that The step of matching the feature sets corresponding to the two frames by optical flow method to obtain the travel distance of the vehicle between the two frames includes: The feature sets corresponding to two consecutive frames are matched using the optical flow method to obtain the travel distance of the vehicle between the two consecutive frames.

5. The method according to claim 1, characterized in that After extracting features from the input images of the multiple cameras corresponding to the multiple frames to obtain feature sets corresponding to the multiple frames, the method further includes: According to the feature set corresponding to each frame, the camera features in each frame are matched in pairs to obtain the matching result corresponding to each frame.

6. The method according to claim 5, characterized in that Before determining the camera calibration result according to the camera pose, the method further includes: Using the matching results corresponding to each frame, adjusting the camera pose to obtain an adjusted camera pose; The determining of the camera calibration result according to the camera pose comprises: A camera calibration result is determined according to the adjusted camera pose.

7. The method according to claim 3, characterized in that The optical flow matching is performed on the feature sets corresponding to the two frames to obtain the travel distance of the vehicle between the two frames, including: Acquire vehicle wheel speed device signals corresponding to the two frames respectively; Performing optical flow matching on the feature sets corresponding to the two frames respectively, to obtain optical flow matching results corresponding to the two frames; The travel distance of the vehicle between the two frames is determined according to the vehicle wheel speed signal corresponding to the two frames and the optical flow matching results corresponding to the two frames.

8. A vehicle-mounted multi-camera joint calibration device, characterized in that: include: An acquisition module, used to acquire input images of multiple cameras corresponding to multiple frames; A feature extraction module, used to extract features from input images of multiple cameras corresponding to the multiple frames, to obtain feature sets corresponding to the multiple frames; A first determination module, configured to determine a plurality of travel distances of vehicles between two frames according to feature sets corresponding to the plurality of frames respectively; A calculation module, used for calculating the camera pose using the travel distance of the multiple vehicles between two frames; The second determination module is used to determine the camera calibration result according to the camera posture.

9. The device according to claim 8, characterized in that The feature extraction module comprises: The feature extraction unit is used to use the Transformer model to extract features from the input images of the multiple cameras corresponding to the multiple frames, so as to obtain feature sets corresponding to the multiple frames.

10. The device according to claim 8, characterized in that The first determining module includes: The matching unit is used to perform optical flow matching on the feature sets corresponding to the two frames to obtain the travel distance of the vehicle between the two frames; repeat the step of performing optical flow matching on the feature sets corresponding to the two frames to obtain the travel distance of the vehicle between the two frames, and obtain multiple travel distances of the vehicle between the two frames.

Images