A Parameter-optimized Camera Response Curve Calibration Method

Abstract

The invention relates to a parameter-optimized camera response curve calibration method, which is characterized in that: the camera is fixed on a tripod bracket, the camera is connected with a computer through a cable, and the exposure parameters of the camera are controlled by the computer to collect image information of a static scene ;Optimize the solution objective function of the response curve, which can effectively reduce the solution redundancy problem caused by useless parameters involved in the process of solving the camera response curve; Solve the solution deviation caused by too many parameters in the traditional process of solving the camera response curve global optimum.

Description

technical field [0001] The invention relates to a parameter-optimized camera response curve calibration method, which belongs to the technical field of computer image processing. Background technique [0002] With the rapid development of computer vision, the application range of digital cameras is also becoming wider and wider. However, due to the limited range of image precision expression of digital cameras, it is impossible to objectively describe the brightness range of the scene. Therefore, in the process of computer vision application, it is usually necessary to use a mapping function to maximize the limitation of the brightness range of the image captured by the camera, and this mapping function is usually called the response curve of the camera. The response curve of the camera can obtain the correspondence between the radiance of the real scene and the acquired image brightness. Therefore, in the high dynamic range imaging technology, it is basically necessary to c...

AI Technical Summary

Problems solved by technology

[0003] The scene information collected by the digital camera forms a two-dimensional array with pixels as the unit through the camera's own imaging array. However, there are a series of nonlinear interference factors in the imaging process of the camera.

Therefore, in order to obtain this nonlinear response mapping relationship, Mitsunaga et al. proposed to use mathematical polynomials to approximately describe this nonlinear mapping relationship. However, the accuracy of the camera response curve obtained by calibration in this way is limited.

However, Mann S et al. strictly limited the nonlinear response mapping relationship, and used the regression method to calibrate and solve it; currently, when calibrating the nonlinear response function of the camera, it usually uses multiple exposure acquisitions for the same scene, and then To solve the response function of the camera, however, in this process, there is a redundant problem of calibration solution parameters, that is, useless parameters in the camera response function also participate in the calibration solution process, so the present invention optimizes this problem Processing, reducing useless parameters in camera response function calibration solution

Images