Design Of Solarcell Defect Detection Sysytem Based On Embedded System

The quality monitoring system in process of solar cell’s production is crucial to the development of photovoltaic industry. In view of the situation that most machine vision quality monitoring systems use PC as their processing control center at present, this paper proposes a solar cells’ defect detecting system based on embedded system to improve system’s performance in cost, efficiency, power consumption and other aspects. The work has been done is listed as follows:(1) The system framework is based on TI’s DaVinci chip TMS320DM6467. The core processors use the embedded processors ARM926 and C6000 series DSP. ARM manages system’s resources and peripherals, while DSP running the algorithm to process the image data. The collaboration of the dual core system use Codec Engine, and communication tool is DSP Link.(2) Based on The DifferenceDmage of Gray and The Gray Level Distribution Image to design pre judgment mechanism. This judgment mechanism is not a accurate detection, just through this mechanism to eliminate the most obvious objects which can be sure without defects, to reduce calculation and improve detection rate.(3) After the pre detect, only these image that may have defects are left, and using deep inspection on these images. The deep inspection algorithm is more accurate.of couse, take longer.(4) In order to implement the detect algorithm on DSP, encapsulate the algorithm according to the xDAIS standard and xDM standard,and transplanted it into TMS320DM6467 platform.at last, compose applications to manage the memory system and 10 interfaces.