Research On Micro Measurement Technology Based On Computer Vision

With sharply development of the scientific technology, traditional measuring method can not satisfy the needs of people. Non-contact measurement method based on computer vision is born at the moment. Computer vision and mechanical automation has been used to do micro operating and measurement etc., the problem on non-contact automation measurement can be solved more effectively. Those image measuring methods can be applied to solve the problems which are not easy to operate, such as the operation of visceral surgery, processing and assembling on precision instruments and so on. This article focuses on the optical microscope imaging characteristics, puts forward an angle calibration on object plane, an automatic Depth-of-Field calibration method and an image height out of the Depth-of-Field measurement method. There are three aspects below:1. Based on the imaging characteristics of optical microscope, a plane tilt calibration method is proposed. The proposed method use a rotating disk, a goniometer and a calibration board with some features are needed, the tilt angle between the object plane and image plane and their geometric relationship can be estimated by the space model. Moreover, it has been proved that the proposed method is quite efficient through experimental verification. It can also contribute to three dimensional applications in computer vision of microsystems.2. As the small Depth-of-Field of optical microscope, an optical microscope automatic calibration method with LBP and active contour model is proposed to measure Depth-of-Field. Besides, it’s an adaptive algorithm, which is able to calibrate the Depth-of-Field according to the roundness and area of dispersion circle under different focus of optical lens. Experimental results demonstrate that the proposed method can be used to measure the Depth-of-Field on optical microscope, which has important significance for the follow-up image processing.3. A super Depth-of-Field height measurement method is proposed to measure object height on an optical stereoscopic microscope to cope with shallow Depth-of-Field. The method uses the principle of binocular vision to construct a measurement model, from which the height of object outside the Depth-of-Field can be obtained. Experimental results demonstrate that the measurement error in the directions Z is 1.6%-7.8%, which indicates that the proposed method can be used to measure a super Depth-of-Field height of micro object within tolerance error.