Sub-pixel-level distortion correction method and apparatus

Abstract

The invention discloses a sub-pixel-level distortion correction method and apparatus. According to the method and the apparatus, display points on a display (4) are controlled to be turned on from a known point by adopting an industrial personal computer, and a camera (1) is controlled to perform image extraction and processing. By adopting the method and the apparatus, the distortion calibration precision can be effectively improved, the sub-pixel-level distortion correction effect can be achieved, the edge calibration effect can be greatly improved, and a visual product can be applied to a more accurate occasion. By applying the method and the apparatus to a projection device, the projection distortion can be remarkably reduced, the projection effect can be improved, and the optical cost can be reduced. An algorithm is simple, and the image processing speed is not influenced, so that the method and the apparatus are very suitable for the fields of high-precision machine vision, camera imaging distortion correction, projection distortion correction and the like.

Description

technical field [0001] The present invention relates to the technical field of image processing, in particular to a sub-pixel level distortion correction method and device Background technique [0002] Imaging equipment on the market, such as cameras, cameras, etc., and imaging products that use lens groups will inevitably produce some distortion, especially in the case of large field of view and short focal length. The effect of distortion is more obvious. Both lens optimization and image algorithms can achieve better distortion correction. Since lens optimization has limited effect at short focus, and involves high technical difficulty and high cost, image algorithm correction is usually used. [0003] At present, the commonly used correction algorithm in the world is Zhang Zhengyou’s calibration method, that is, through the standard black and white checkerboard image, the image position of each corner point is calculated, and compared with the theoretical image position, ...

AI Technical Summary

Problems solved by technology

The influence of this error is not obvious on ordinary video and projection, but it has a greater impact on more precise applications or ultra-short focal lengths with relatively large distortion.

[0004] For example, in the field of binocular stereo vision, the distortion of the two cameras is required to be completely corrected to match the image. Pixel-level distortion will directly reduce the matching accuracy and affect the matching quality.

In the field of high-precision visual dimension detection, distortion will directly cause measurement errors and affect measurement accuracy

[0005] In addition, the algorithm cannot correct the distortion of projection products

Images