Automatic IC Character Recognition System for IC Test Handler Based on SVM

An automatic IC character recognition system for IC test handler is proposed in this paper based on support vector machine (SVM). It can achieve character recognition on the chip surface. Firstly, the chips' images are captured and binarized using Otsu algorithm. And median filter is used to remove the images' noise. Secondly, character segmentation is done to get the image of each character through projection feature. Character images are normalized in order to convert to same size. Thirdly, feature extraction is performed based on grid feature. Fourthly, the classifier based on SVM is used for sample training and character recognition. Finally, the character recognition system based on SVM is built. And experiment on character recognition for IC chips is performed. The experiment results demonstrate the effectiveness and feasibility of the proposed system.