High-speed CMOS Camera Driver Design And Optical Image Preprocessing

Large-field image processing systems are widely used in the field of scientific research and life.With the development of technology,the speed of image transmission is greatly improved.At the same time,with the increasing demand for application,the demand for imaging quality of imaging system is becoming higher and higher.With the enlargement of the field of view,the distortion of the image becomes more and more serious when using the large field of view lens to obtain the image.Therefore,this paper designs a large field of view research camera based on FPGA combined with the actual project requirements.Camera design includes two aspects: Image acquisition and transmission and image data preprocessing.In this paper,the image sensor CMV4000 of CMOSIS Company is used as the detector,the VIRTEX5 FPGA of XILINX Company is used as the processing platform,and the CameraLink interface is used as the output channel of the camera to realize the function of high-speed image acquisition and transmission.In the case of the resolution of 512 × 512,the frame frequency of the camera is up to 200 fps,which meets the application requirements in most cases.In this paper,the multi-channel output CMOS sensor is used as the photosensitive component of the system.Due to the structure of the CMOS image sensor,the non-uniformity of the sensor's light response is inevitable.The spatial stripe noise appears in the collected image,which reduces the quality of the image.In this paper,the non-uniformity correction algorithm is studied,and the non-uniformity of the sensor is calibrated by the method of calibration combined with the characteristics of the CMV4000 sensor.The experimental results show that the spatial fringe noise of the CMV4000 image sensor is effectively eliminated whichimproved the quality of imaging.At the same time,in order to obtain the large field of view image information,the camera uses 90 degrees wide angle lens,the image has serious distortion.In this paper,the correction algorithm of wide-angle lens distortion is studied.For the lens used in this paper,a correction algorithm is proposed,which is easy to be realized in real time on FPGA.The experimental results show that the distortion rate of the image is obviously reduced.