Fault Identification in Printed Circuit Board Using Thermal Image

The different defects in PCB (Printed Circuit Board) are hairline, pin-hole, wrong size hole, breakout, open circuit, mis-contact, spur, mission feature, etc, and it is essential to detect these defects effectively. The Conventional automatic test equipment (ATE) offers some limitations in testing the produced PCB which has been attached, inserted and soldered the electronic components. The no contact problem, rapid image acquisition, easy operation, and simple testing reconfiguration have the advantages of the thermal imaging diagnostic system. Therefore, this approach has been widely applied in the past decade to diagnose faults on PCB.
To analyze the thermal image of PCB for faults detection proposed a novel vector quantization (VQ) based approach, which can reduce the memory size, for thermal image analysis of PCB diagnosis. Appling the proposed VQ-based approach, the gold thermal image is coded into a codebook and compared with the board under test (BUT) to identify the image blocks with faults, instead of the whole thermal image. This proposed method is demonstrated to be very effective in abnormal functional block identification for PCB based on thermal image.