Image correction processing method and device

Abstract

The embodiment of the invention provides an image correction processing method and device, and the method comprises the steps: calculating an image view angle corresponding to each image point in a to-be-processed image, and correcting image points, with the image view angles being smaller than or equal to a preset view angle, in the to-be-processed image through a first distortion sub-model; according to the invention, correcting the image points with the image visual angles greater than the preset visual angle by using the second distortion sub-model, so that the image points with differentdistortion degrees can be accurately corrected by using different distortion models, and the problem of inaccurate correction caused by correction of the image points by using the same internal reference distortion model is effectively reduced; and applications such as distance measurement or detection of visual tasks are facilitated.

Description

technical field [0001] The present invention relates to the technical field of cameras, in particular to an image correction processing method and device. Background technique [0002] Vision-based target perception is an important part of autonomous driving environment perception, among which camera internal parameter calibration is the most basic part of visual perception. [0003] However, the existing cameras all use a unified internal reference distortion model for internal reference calibration. Therefore, when the camera performs distortion correction processing on the image points of the captured images, whether it is an image that is closer to the central area of ​​the camera lens in the image (image Distortion is small) or the image closer to the center of the camera (larger image distortion) is used to correct the distortion of the image with the same internal reference model, resulting in over-correction or under-correction of the image, which affects the subsequ...

AI Technical Summary

Problems solved by technology

[0003] However, the existing cameras all use a unified internal reference distortion model for internal reference calibration. Therefore, when the camera performs distortion correction processing on the image points of the captured images, whether it is an image that is closer to the central area of ​​the camera lens in the image (image Distortion is small) or the image closer to the center of the camera (larger image distortion) is used to correct the distortion of the image with the same internal reference model, resulting in over-correction or under-correction of the image, which affects the subsequent use of the image to target Applications such as detection or ranging

Images