Design Of High Resolution Image Acquisition Device Based On FPGA

Image acquisition and processing technology has been applied to all aspects of production and life.In response to the demand for high-resolution and high-definition imaging in fields such as machine vision and artificial intelligence,a high-resolution image acquisition device based on FPGA has been designed.Real time processing,parsing,and transmission of image data through FPGA chips,Utilizing the advantages of FPGA chips in real-time image processing,achieving 20 million pixels(5472×3648)for high-resolution image acquisition and display.On this basis,the Bilateral filter algorithm is used to filter the image noise that may appear during the transmission of the image acquisition device,which further improves the image clarity.The main research content is as follows: 1)Based on the design scheme,compare and analyze the key chips of each module of the image acquisition device to complete the chip selection of each module.Combined with the chip data manual,complete the hardware circuit design.2)Using a top-down design approach,the software design of the image acquisition device is completed according to module functions,including power initialization,register configuration,LVDS image data transmission and analysis,and DDR3 SDRAM read and write cache logic design.By designing a reasonable cache logic,the problem of image tearing caused by inconsistent read and write rates of image data has been solved.3)The Bilateral filter algorithm is implemented on FPGA by using look-up table and parallel pipeline processing,and the portability of Bilateral filter system is improved by modular design and parametric programming.Finally,the designed image acquisition device is tested and validated,and the experimental results show that the image acquisition device can perform 20 million pixels(5472×3648),5.95 fps frame rate display.The power consumption of the image acquisition device before adding the bilateral filtering algorithm is 3.12 W,and after adding the bilateral filtering algorithm,the power consumption is 3.60 W.The filtering effect of the bilateral filtering algorithm is evaluated by the PSNR(Peak Signal to Noise Ratio)and SSIM(Structural Similarity)values before and after filtering.The results show that the PSNR increases from 21.8277 d B to 25.1119 d B,an increase of 3.2842 d B;The SSIM increased from 0.4840 to 0.5896,an increase of 21.82%,and the image quality was enhanced.