LED芯片自动分选系统中的视觉定标问题研究

Study of Vision Calibration in LED Chip Auto-sorting System

本文针对LED芯片分选系统的成像系统的特殊条件及较高的精度要求,进行了视觉定标模型的计算和构建。根据共面约束进行简化,推导出了新的带有两种一阶畸变的非线性视觉定标模型。提出了多段式"求解范围—优化"的模型求解方法,并针对模型的非线性优化问题,设计了相应的遗传算法优化解法。并设计了适用于LED芯片分选系统的实验方法。实验结果完全满足LED芯片分选系统的精度要求。本文所提出的方法,对于其它的摄像系统模型参数求解也具有很大的意义1。

For the special condition and high precision requirement of the LED chip sorting system,the model of vision calibration is built and calculated.Through the predigestion of eoplanarity restriction,a new non-linear vision calibration model which has two firstorder aberrations is aqiured.And a Subsection ＂range solue- optimize ＂solution is put forward. This article designs corresponding Genetic algorithm for the optimize problem.h also design test measure suit to LED chip sorting system.The test results meet the precision requirement of system. The method put forward by this paper is suitable for parameter solution of other camera system model.