Error analysis and improved calibration algorithm for LED chip localization system based on visual feedback

Positioning accuracy of chip directly impacts on the quality and efficiency of LED chip production. The purpose of this paper is to improve the chip positioning accuracy by improving the camera calibration algorithm of LED chip visual positioning system. Firstly, by making the error analysis for the visual positioning system, the systematic errors of each parts of the system are obtained, and the relationship between chip positioning error and chip position distribution in image is found. Then, according to the result of error analysis and the characteristics of the chip positioning process, an improved calibration algorithm is proposed to improve the chip positioning accuracy. This improved algorithm solves the calibration parameters in two steps, which highlights the main cause of errors in calibration process and meets the requirements of chip positioning. Finally, the experiment results show that the proposed algorithm can improve the chip positioning accuracy obviously, and has good stability and robustness.