Research On Image Automatic Identification Technology To Valve Small End-face Pore Disfigurement

In the past twenty years, the research of the technology of image automatic identification rose healthily. The applications of this technology in many fields are more mature nowadays. The goal of this paper arousing the research task of image automatic identification technology on valve small end - face pore disfigurement, is to resolve a practice issue on how to check small end-face pore disfigurement in the industry produce process, using the technologies of image automatic identification and advanced control.This paper, firstly, makes a statistical comparison on the character of valve small end-face disfigurement, and obtains the rules of how the disfigurement distribute and the rules of what the figures are. Then we take many photos on the valve small end-face under special setting. We analyse and compare these images, dig out the image difference on regular and irregular valve, and so provide the design base of development of measure device on the valve small end-face disfigurement on base of image automatic identification.The paper design a suit of experiment device, including image real time collection system and image disfigurement identification system, to collect image and identify the disfigurement in the image. The image real time collection system are composed of lamp-house, vidicon, image collection card , PC and the image collection program running on PC. Image disfigurement identification system makes use of image pre-process arithmetic and image identification arithmetic to develop an image identification program. The image collection program is developed on PC by VC based on the image collection card,which can take the tasks of image collection real time and can adjust the image quality by changing the brightness and contrast reasonably. Image disfigurement identification program designs identification arithmetic by using C/C++ program language, and counts the numeral criterion of the regular & irregular valves , and then draws a conclusion.By this experiment device, the paper makes a practicality test to some valves. The content are: (1) MIX-TEST. We mix the regular and irregular valves into a set of sample, and check the potential disfigurement in this sample. (2) REPETITION-TEST. We take four different images of each sample from different place and different azimuth, and check the potential disfigurement. This experiment shows the main targets of the device, for example, precision and veracity, achieve to a satisfied level. So what the paper arouses and how to resolve the issue are completely reasonable and significative.Finally, we make a good expectation about a practical measure device from our experiment one and some wonderful suggestion is also given. We hope it would be finished in the future.