Design Of CCD Color Sorter Data Acquisition And Sorting Drive System Based On FPGA

The CCD color sorter acquired images by CCD sensors, analyzed images by microprocessors and sorted the material automatically by the solenoid valve which drove by the FPGA with advantages of efficiency, safety, automation and no damage to the material, the CCD color sorter was widely used in agricultural products processing and industrial inspection.Image acquisition and material sorting were studied in this paper, an image acquisition and solenoid valve driving system controlled by the FPGA was designed accordingly. In this system, the FPGA chip acquired image data through Camera Link interface, controlled external FIFO chip for ping-pong storage and then transported image data to the DSP. A solenoid valve driving board was applied to receive the DSP’s sorting signal and switch the solenoid valve in the system’s sorting part.Firstly, 6 kind of circuits,which include FIFO storage circuit,MOSFET driving circuit, FPGA minimum system circuit,communication interface circuit,power circuit and Camera Link interface circuit, were designed and these circuits were proved to be met the requirement of image acquisition and material sorting by both simulations and experiments.Secondly, a driving strategy of high voltage opening and low voltage PWM maintaining was selected to respond the solenoid valve rapidly and control accurately after a compare of three common solenoid valve driving strategies. As the main control chip,a suitable FPGA was utilized and a custom protocol based on RS-485 was designed according to the system’s requirements.System’s software design was developed in Verilog HDL and in a modularization and top-down method. In the image acquisition part, camera controlling module, image acquisition and storage module, data transmission module to the DSP were implemented, In the material sorting part, baud rate clock module, bytes transmit/receive module, frame transmit/receive module, CRC checking module, solenoid valve drive module and est module were implemented. These modules’ correctness was verified by Modelsim 10.0c simulation and SignalTap test.The online debugging of the whole system was adopted to observe whether the image acquisition part and material sorting part could meet the design acquirement. The following conclusions could be drawn after the online debugging:(1) image data storage and transmission at 40 Mhz clock and camera controlling was implemented.(2) the communication which utilized the custom protocol based on RS-485 with the DSP was of stability. The high voltage opening low voltage maintaining strategy improved the opening speed and reduced the energy consumption of the solenoid valve. The precision of injection could also be improved by modifying the solenoid valve’s parameters.