Feature extraction and matching location algorithm based on wavelet moment and curvelet transform

In the process of wire bonding, pattern recognition system is often applied to measure the deviation of the actual chip relative to the sampling position (eye-point) in order to ensure the location accuracy and speed. In this paper, a new feature extraction and matching location algorithm based on wavelet moment and curvelet transform is proposed for wire bonding. The algorithm combines the good properties of wavelet moment and curvelet transform, in addition, the application of coarse and fine searching improves the matching efficiency and accuracy. Euclidean distance is chosen to finish the matching location of the eye-point images and images to be matched. Experimental results show that the algorithm has high performance in the positioning accuracy and detection efficiency. The proposed algorithm is also robust and suitable for automatic wire bonding vision positioning system.