Dsp-based Vessel Detection System Research

Aimed at solving the key problem in food and medicine industry an image detection system based on machine vision was proposed. The processing was divided into three steps: firstly digital images of glass container collected, secondly the obtained digital images processed and finally recognized. Images were collected by using CCD camera. The digital-images were treated by means of processing system. The image processing system mainly consisted of DSP and FPGA. It is programmable and of high calculation speed.The original gray scale images were filtered by filter based on mathematics morphology and the edge was detected by means of so called extremum based on edge detection algorithm.An real-time image processing system was presented against the need of online detection. A novel filtering method based on mathematics morphology was used in image processing which has better performance than commonly used algorithms . An extremum based on edge detection algorithm was proposed. The extracted edge obtained by means of this algorithm is the true edge and it has good continuity.