Panoramic camera and shooting method thereof

Abstract

The invention discloses a panoramic camera and a shooting method thereof. The panoramic camera of the invention comprises: N sub-cameras; a synchronous control module; a splicing and correction module, which is used to cache m rows of continuous pixel values ​​in a captured image in a sliding window manner, And according to the preset correction table and the cached m rows of pixel values, complete the correction of the pixel values, and sequentially output the corrected pixel values ​​of each row, wherein, m is greater than the preset maximum distortion row of the image number+2, and m is less than the total number of lines of the captured image; the splicing and fusion module is used to synchronously capture the respective sub-cameras under the control of the synchronization control module according to the overlapping parts of the captured images of the respective sub-cameras The captured images are stitched into a stitched image; and a transmission module. The panoramic camera of the invention has better processing efficiency and better processing quality, and is more conducive to miniaturization of system design.

Description

technical field [0001] The invention relates to the field of video monitoring, in particular to the shooting technology of a panoramic camera. Background technique [0002] In order to achieve full coverage of key monitoring areas, multiple cameras are generally used for shooting. For example, for the video monitoring of the intersection of a city's main road, multiple network cameras are installed and arranged from all directions, so as to form a 360-degree monitoring without dead ends. [0003] However, in the above-mentioned technical solution, each network camera independently transmits the video data to the monitoring center and displays them independently on the large screen of the monitoring center. This is not only costly, but also the independent picture scene brings the viewer visual Experience is also greatly affected. [0004] In order to solve the above problems, the real-time video collected by multiple front-end network cameras is usually sent to the central...

AI Technical Summary

Problems solved by technology

However, since each network camera collects video images and transmits them through the network, it will cause different delays

Due to different time delays, there may be mismatches during image stitching, resulting in dislocation or overlapping of moving objects in the composite image

[0005] On the other hand, although there is also a solution to obtain a panoramic image by directly splicing the real-time video collected by multiple front-end network cameras or multiple cameras of the network camera at the front end, the consumption of system resources is large and the splicing effect is not good. The problem

Images