Automatic classification of SMD packages using neural network

This paper proposes a surface mount device (SMD) classification method for printed circuit board (PCB) assembly inspection machines. The package types of SMD components must be classified to create a job program for the inspection machine. In order to reduce the creation time of the job program, we developed an automatic classification algorithm for SMD packages. We identify the chip-type packages by color and edge distribution of images. The input images are transformed to the hue saturation intensity (HSI) color model, and the binarized histograms are extracted for the H and S spaces. Also the edges are extracted from the binarized image, and quantized histograms are obtained for the horizontal and vertical directions. A neural network is then applied to classify package types from the histogram inputs. Experimental results verify the usefulness of the proposed method.