Study On Intelligent Detection Of Geometric Parameters Of Chip

In this thesis, an image processing and analyzing system based on binocular visual model was established, and a corresponding image detecting software for determination of geometric parameters of chip was developed. Compared with the existed detecting algorithm based on monocular visual system, the detecting algorithm put forward in this thesis shows higher accuracy and stronger robustness. Experiment results show that this system is effective in off-line detection of the chip geometric parameters.The main work done in this thesis is as follows:1.The present detection technique for geometric parameters of chip was discussed in detail. The excellence of detection with computer vision was summarized.2. Based on analyse of the frame of binocular visual model, the hardware of this system was designed. With this system, the cutting process can be recorded accurately.3. According to stereovision theory, the detecting algorithm was completed with a series of image processing methods and numerical methods. With this algorithm, the chip geometric parameters can be calculated exactly.