An intelligent vision system for inspection of packaged ICs
D.P. Mital, Teoh Eam Khwang · 2003
A visual inspection system for recognizing defects in IC packages is presented. In this system, the iconic domain of two-dimensional gray level image analysis based on the feature extraction method is used. A software package is developed to carry out the extraction of defects that may exist on the IC leads and the surface of the IC chip. The recognition of these defects is achieved by comparing the image data with a predefined threshold value. To improve the speed of performance of the system, an auto-windowing routine is incorporated in the system to map out the relevant areas for image processing. The inspection algorithms and various processes used by the system are outlined. From the experimental results obtained, it is shown that high accuracy and good reliability can be achieved by the proposed system.>