Large vehicle panoramic splicing method

Abstract

The invention relates to the field of image data processing, in particular to a large vehicle panoramic stitching method, which comprises the following steps: acquiring video data of at least one camera, and preprocessing based on the type of the camera to obtain a video frame; obtaining processing data based on image feature point matching of the video frame; horizontally splicing the processed data; and outputting the spliced image. One or more cameras are arranged so that all data of the appearance of a large vehicle can be obtained, a wide-angle camera is generally adopted for one camera, and a flat-angle camera is generally adopted for multiple cameras, so that corresponding video images can be matched to obtain processing data of a video frame with a complete section of the vehicle, and then horizontal splicing is carried out to obtain the complete appearance of the vehicle, so that the vehicle can be better shot and processed, and the measurement and detection efficiency is improved.

Description

Technical field

[0001] The present invention relates to the field of image data processing, and more particularly to a large vehicle panoramic splicing method. Background technique

[0002] Due to the bayonet space limit, the use of the monitoring camera cannot take a full large vehicle, generating a large vehicle panorama into the key to the bayonet monitoring.

[0003] In the face of problems such as high vehicle driving speed, traffic jams, real-time requirements, etc., existing image splicing algorithms have shortcomings in splicing effects and time consuming. Inventive content

[0004] It is an object of the present invention to provide a panoramic splicing method of a large vehicle, aimed at solving the problem of the vehicle splicing method.

[0005] In order to achieve the above object, the present invention provides a large vehicle panoramic splicing method, including: acquiring video data of at least one camera and performs a video frame based on the camera type; [00...

Examples

Embodiment 1

[0051] S201 acquires video data of at least one camera and pre-processed to get a video frame based on the camera type, the camera being a wide angle camera;

[0052] The video data generated by the wide-angle camera will produce distortion.

[0053] S202 matches the image feature point of the video frame to obtain processing data;

[0054] The specific method is to deform data on video data. The image captured by the wide-angle lens camera often has a problem of barrel distortion. The reason is that the wide-angle lens uses a convex lens, junior high school physics knowledge tells us that the convex lens will give a convergence of the light, which is caused by the refraction of light. The farther away from the center of the lens, the stronger the refractive effect, and the photos taken will be rounded in the center of the lens, and the distortion is expanded. Therefore, the wide-angle distortion modification algorithm can be used to process video data, resulting in normal video d...

Embodiment 2

[0066]The S401 acquires video data of at least one camera and pre-processes the video frame based on the camera type, the camera being two flat angle cameras;

[0067] By setting two flat angle cameras, the upper portion and the lower portion of the vehicle are filmed to obtain the corresponding video data.

[0068] The specific steps are:

[0069] The characteristic match of video data based on two flat angle cameras determines the up and down offset distance of the two flat angle cameras; the video data of the two graphic cameras is spliced ​​based on the up and down offset distance to obtain the processing data S403 horizontally splicing the processing data.

[0070] Since the position of the two square angle cameras is relatively fixed, each video frame will have the same upper and lower offset distance. It is therefore necessary to characterize the vehicle in the two video data to determine the up and down offset distance of the splicing, so that video data that splices the u...

Technical field

[0001] The present invention relates to the field of image data processing, and more particularly to a large vehicle panoramic splicing method. Background technique

[0002] Due to the bayonet space limit, the use of the monitoring camera cannot take a full large vehicle, generating a large vehicle panorama into the key to the bayonet monitoring.

[0003] In the face of problems such as high vehicle driving speed, traffic jams, real-time requirements, etc., existing image splicing algorithms have shortcomings in splicing effects and time consuming. Inventive content

[0004] It is an object of the present invention to provide a panoramic splicing method of a large vehicle, aimed at solving the problem of the vehicle splicing method.

[0005] In order to achieve the above object, the present invention provides a large vehicle panoramic splicing method, including: acquiring video data of at least one camera and performs a video frame based on the camera type; [00...

Examples

Embodiment 1

[0051] S201 acquires video data of at least one camera and pre-processed to get a video frame based on the camera type, the camera being a wide angle camera;

[0052] The video data generated by the wide-angle camera will produce distortion.

[0053] S202 matches the image feature point of the video frame to obtain processing data;

[0054] The specific method is to deform data on video data. The image captured by the wide-angle lens camera often has a problem of barrel distortion. The reason is that the wide-angle lens uses a convex lens, junior high school physics knowledge tells us that the convex lens will give a convergence of the light, which is caused by the refraction of light. The farther away from the center of the lens, the stronger the refractive effect, and the photos taken will be rounded in the center of the lens, and the distortion is expanded. Therefore, the wide-angle distortion modification algorithm can be used to process video data, resulting in normal video d...

Embodiment 2

[0066]The S401 acquires video data of at least one camera and pre-processes the video frame based on the camera type, the camera being two flat angle cameras;

[0067] By setting two flat angle cameras, the upper portion and the lower portion of the vehicle are filmed to obtain the corresponding video data.

[0068] The specific steps are:

[0069] The characteristic match of video data based on two flat angle cameras determines the up and down offset distance of the two flat angle cameras; the video data of the two graphic cameras is spliced ​​based on the up and down offset distance to obtain the processing data S403 horizontally splicing the processing data.

[0070] Since the position of the two square angle cameras is relatively fixed, each video frame will have the same upper and lower offset distance. It is therefore necessary to characterize the vehicle in the two video data to determine the up and down offset distance of the splicing, so that video data that splices the u...

Images