Verification of error in SMB manufacturing still depends on human visual inspection. With high density, small chip size and high complexity of the surface mount board, an automated smart inspection system is required to reduce manufacturing flaws. Above all, the system should be capable of linking errors to their causes in the manufacturing process, so that restorative steps can be taken quickly. The typical flaws in SMB manufacturing are missing chips, correctness of chips, alignment and positioning of chips, and quality of solder joints. In this report, a solid smart vision system architecture and several algorithms are created to deal with this demanding and difficult problem…..
Because of the absence of an industrial standard, several marking schemes appear on the same chip from different manufacturers, and for a given manufacturer, different formats are employed on different packages. The issues of printing quality, location variation and format inconsistency along with handling in the manufacturing process make the chip image quality to be less than perfect. The emphasis of high speed at the cost of resolution makes character recognition and dimensional measurement almost impossible. Furthermore, the solder joint visual images captured under the poor resolution and varying lighting conditions are insufficient in details that are required for fault categorization. To verify the chip at high-speed, the entire chip image is captured and processed at a time. Several algorithms are produced to take care of resolution difference among different chip sizes that can have a factor of 10-15. To get rid of image imperfection and for exact chip identification, a chip reference database is created. Information on the board under test and its chips is directly downloaded to the vision system. An extensive graphical user interface with the electronic device and checking board database storage system is designed for quick access and update of information. To cope with varying character and symbol sizes, an iterative sampling technique is utilized to separate characters and symbols from the background and subsequently bounded by a rectangular array…..
Source: City University of Hong Kong