Development Of Object Identification And Control System Of Color Sorter

China is a big agricultural country, with its grain output ranked first in the world and increased year by year. The rapid development of grain processing industry comes with the improvement of people's living standard in recent years. Color sorting machine plays a key role in the process of grain sorting, as one of the important equipment in the grain processing industry. Currently available CCD color sorting machines on the market, have excellent performance with expensive cost. Reducing the cost of the color sorting machine to make economic profits is an important issue. The purpose of this paper is to develop a cheap target recognition and control system of color sorting machine based on CIS, achieving the function of grain sorting with the image capture through CIS and processing the image with algorithms then output control signals.A new design of target recognition and control system of color sorting machine, which captures images with a CIS contact image sensor, and processes images with DSP processor, is put forward in this paper. This design programs the Verilog files with Quartus II of Altera and simulate them with Modelsim. Driving the CIS and ADC to capture and transmit images with configuring the FPGA. A data transfer parallel interface between FPGA and DSP is designed to read the image data rapidly. The simulation experiment based on MATLAB about the image processing, implemented by processing images with appropriate algorithm, and outputting sorting and control signals.Compared with the traditional CCD color sorting machine, the accuracy of the system based on CIS is comparable with CCD color selector. This system generates relatively clear images with cheap cost and convenience in assembly, because of the application of CIS. Programmable logic device is used to driving the CIS and ADC, making the digital circuit integrated with a high degree, and timing control becomes simple, stable and easy to upgrade. Combined with DSP processor, the image processing algorithm is able to accurately identify the heterochromatic grain position information, meeting the designed accuracy and instantaneity.