With the rapid development of the automotive industry, aluminum wheel as an integral part of the car, has a broad market prospects.Aluminum wheel is far beyond the traditional steel wheel in the quality, safety, performance and other aspects, and the diversity of its appearance,high cost performance has become the consumer of choice.But due to problems of production process, on the aluminum wheel hub surface exists the phenomenon of sand holes, so aluminum wheel needs quality and safety testing.At present,the wet tightness detection method is widely used at home and abroad,this method is generated by artificial to identify whether there are air bubbles, leading to low efficiency, and it is not conducive to automated testing.While wet tightness testing method require a higher capacity for the staff, for the continuous generation of visible bubbles, are easy to determine the presence of an aluminum wheel surface blisters,but the tiny sand holes and interference bubble is difficult to distinguish, the bubble is prone to misjudge.Based on aluminium wheel as the research object, adopting the high precision CCD camera replace the artificial judgment way, through the image collection and processing system, the CCD camera in receiving and processing of image data, and processed data to the computer for display and judgment, so as to solve the artificial misjudgment, low production efficiency and unable to automated testing, etc.Most image processing method is given priority to with the software, but when the common computer processor to realize complicated algorithm, will cause the loss of image quality, low frame rate,and the processing speed is low.Late for high speed and convenient real-time system upgrades, flexibility, this paper presents a method of image processing based on FPGA.This paper introduce the working principle of Camera link interface and PCI bus in detail,design a set of digital Camera image collection and processing system based on Camera link+FPGA+PCI. With FPGA as the core of image acquisition and processing system, combined with high-precision digital camera CCD signal acquisition,using fast median filtering for image preprocessing,Sobel edge detection algorithm is proposed by using System Generator software design,and finally mapped to inside the FPGA, so as to realize the processing of video images, through PCI bus upload the processed data to the computer, the software interface of gas tightness detection system was designed on MATLAB, and the results were displayed and judged.To be measured by aluminium wheel rotation, and observe the same position whether bubbles or bubble density methods eliminate interference bubbles, to achieve the purpose of the feature extraction of automobile wheel hub bubbles, which abandoned the traditional way of artificial, realized the online monitoring, monitoring to ensure the accuracy and reliability of the system operation.Upon completion of the hardware circuit and driver circuit design, the fast median filtering, Sobel edge detection algorithm and PCI bus were simulated and tested, and on the development of EDA software to verify the feasibility and stability of the program,and used MATLAB software to develop the GUI air tightness test system. |