| High-speed image acquisition and real-time processing have a significant value in aerospace,military, industrial inspection and other fields. Industrial inspection has raised a high requirementfor image sensor. Linear array CCD is suitable for industrial field environment with high resolution,low cost advantages.Image processing system has new requirements for the speed of image acquisition, processing and transmission in the practical application, requires a high requirement for real-time performance. Field Programmable Gate Array(FPGA) is currently the most widely used one of programmable logic devices. A large number of logic and storage resources exist in FPGA, parallel operation and pipeline can be implemented in design. Compared to ARM and DSP, FPGA has many advantages, such as high-speed processing ability, universal transplantation and can be dynamically configured with flexibility, which are very suitable for image real-time processing.A linear array CCD intelligent image processing system based on FPGA was designed and implemented in this thesis.In the system, the linear array CCD chip TCD1209 which is produced by Toshiba Corporation was adopt as an image sensor, ADI’s special image processing chip AD9945 was used for amplifying and analog/digital (A/D) conversion of CCD’s output signal. By the FPGA chip from Altera’s Cyclone IV series,signals of the driving for analog front end, controlling the peripheral devices and processing the collected images were generated. Modules, like image data acquisition, image data cache, FIFO,USB2.0 and RS485 communication, image filtering and edge detection, had been designed with hardware description language Verilog HDL,and front-end simulation ofentire system had been done with development platform QuartusⅡ 13.0 provided by Altera, as well as simulation tool ModelsimSE 10.0 provided by Mentor Corporation.The test results show that linear array CCD intelligent image processing system based on FPGA can implement high-speed data acquisition, transmission and processing of image signal with the advantages of high integration, excellent real-time performance and universal transplantation. The noise was suppressed and the edges of the image were enhanced. Therefore, the system is more conducive to industrial inspection. |