Research Of Objective Measuring Technique Based On Image Distance Measurement By Defocusing

With the continuous development of modern industrial technology, CCD camera isused to shoot a single image to do non-contact dimensional measurement of the object.However, height difference between the target objects in the optical axis direction willaffect the measurement accuracy of the different sections. In this kind of work piecemeasurement, it is the key to extract effectively object depth information. The stereo andmotion vision method of the passive ranging algorithm both are unable to find thecorresponding feature points matching. So In this paper, such problems are analyzed. Theimage distance measurement by defocusing, which is based on monocular vision, is used toextract the depth information of the target object. Then According to the obtained depthinformation, it is possible to gain high precision dimension measure of the object differentsection.In this paper, three classic basic visual model of defocus ranging algorithm are brieflyintroduced, such as Pentland ranging method, Subbarao ranging method and rangingmethod based on invariant moments, and we point out their limitations on the practicalapplication according to their ranging Algorithm. Then we propose a defocus rangingalgorithm based on shades of gray.First，we use gray gradient method to determine the depth of information between twopoints on the surface. Secondly, we determine the location relationship between the actualimaging surface and focus image plane, according to the size relationship of diffuse spotradius between two things in the imaging surface of defocus imaging model. Then, bymeans of the S-transform, which can be used to estimate the point spread function, thedistance formula between the object and optical lens will be obtained. The proposedalgorithm is improved to complete dimension measure of the object different section, which is based on defocus ranging, by the experimental equipment and analyses ofexperimental errors.