Research On Synchronous Acquisition Of Multi-channel Images And Preprocessing Technique

With the continuous development of 3D technology,3D face modeling has become a research hotspot,and face image acquisition is the basis of 3D face modeling.This thesis is based on the requirement of the 3D modeling equipment proposed by our research team,a multi-channel image synchronization acquisition system based on FPGA is designed,the preprocessing algorithms are investigated and FPGA implementation are accomplished.The main work of this thesis is as follows:(1)A multi-channel image synchronization acquisition system based on FPGA is designed.In this system FPGA is used as the core control,CMOS OV5640 image sensor as the images acquisition,DDR3 SDRAM as the images cache,the Verilog HDL is used to realize the logic design of the modules,such,as acquisition module,buffer module,HDMI display module and Ethernet trarnsmission module.(2)The image denoising,image contrast enhancement and image edge detection algorithms are studied.The FPGA implementation of image fast median filtering,histogram equalization and Sobel edge detection algorithm are completed.(3)In order to automatically detect the anomalies in the image acquisition process,the image black detection and image blur detection algorithms are researched.An image blur detection algorithm based on edge and transform domain combination is proposed.The algorithm can effectively detect the acquired blurred image,which has higher detection accuracy.The system test shows that the designed face image acquisition system can synchronously acquire and display images,and realize the real-time preprocessing.All these make a solid basic for our development of the whole equipment of 3D modeling.