Camera Parameter Calibration Method Based On Minimal Target

In recent years,with the rapid development of science and technology,computer vision technology has become more and more widely used in real life situations such as medicine,archeology,military affairs,and movie animation.People's demand for image processing and a series of technology are increasing.Stereo vision is one of the hot research directions in the field of computer vision.Simultaneously,camera calibration is the foundation and necessary link for realizing stereo vision technology.Therefore,how to achieve a fast,high-precision,robust camera calibration method has important theoretical research significance and practical application value.The essence of the camera calibration process is to establish a mathematical model between the computer image coordinates and the world coordinates in order to obtain the internal and external parameters of the camera.So far,there are many methods for the calibration of traditional cameras,but most of the methods are complicated in calculation process.In addition,the production process of the target is high.There are few ways can overcome the limitations of ambient lighting.Most methods cannot overcome the limitations of ambient lighting and unable to complete the calibration in a dark environment.On the basis of analyzing and summarizing predecessors,this paper proposes a calibration method based on the minimal target,which mainly completes the following work:(1)Determine the calibration scheme.This paper builds on the traditional camera calibration method,adds a geometric model constraint to the pinhole imaging model of the camera,uses the six points of space to establish a minimal target model and deduces the algorithm.The camera's internal and external parameters are solved within the linear operation range,thus simplifying the constraint equation solving process,providing a theoretical basis for follow-up experiments..(2)Designed to produce a minimalist target.According to the feature points involved in the mathematical modeling of the calibration scheme,a minimal target model was created.In addition,in order to improve the defect that the traditional calibration method can not be calibrated in the lack of illumination or dark environment,an active target is fabricated by installing an infrared laser diode to realize the calibration adapted to a low-light environment.By obtaining the target feature points in different light environments,it is proved that the minimal target is suitable for more calibration scenes.(3)Realize the feature point image coordinates extraction.In the process of camera calibration,the extraction of feature point image coordinates directly affects the accuracy of the camera calibration.Therefore,under the premise of analyzing the feature point image in the target image of this paper,a feature point that integrates various image processing steps is realized.The coordinate extraction method can effectively and quickly detect the target image feature point coordinates.(4)Experimental design and results analysis.According to the calibration method of the camera parameters based on the minimal target in this study,the calibration experiment was designed to verify the theory of the method,and compared with the accuracy,complexity,and stability of Zhang Zhengyou's checkerboard calibration method.Both methods can control the error within 0.6 pixels,and achieve a good calibration effect.At the same time,this method reduces the complexity of the algorithm,but the stability needs to be further improved compared to Zhang Zhengyou's checkerboard calibration method.