Research On Image Analysis System For Quantitative Evaluation Of Material Surface Corrosion Feature

Nowadays, corrosion characteristic on the surface of material is mainly evaluated and analyzed through manual observation and analysis in China Industries 59th Institute Natural Environmental Testing Center. Manual observation and analysis on corrosion characteristic is influenced by individual differences and experiences, which is hard to guarantee the reliability of evaluation on the one hand and to meet the quantitative treatment in the information era on the other hand.Digital image has a high resolution, which can reflect and evaluate corrosion feature directly by image processing and analysis. Digital image processing has high accuracy and speed. Digital image processing technology is introduced into evaluation and analysis of corrosion characteristic on the surface of the material, which is called the image analysis system on the surface of corrosion characteristic. In the Image Preprocessing Module, the main target is to program various image enhancement algorithms, which is used to improve image quality, highlight the target of the image and facilitate follow-up image segmentation. In Image Segmentation Module, the main achievement is to eliminate over-segmentation of corrosion characteristic image, extract erosion feature accurately through analyzing the reason of generating over-segmentation in traditional watershed segmentation algorithm, and, region of interest(ROI) segmentation processing and human-computer interaction technology are introduced in order to improve flexibility and adaptability of image segmentation. In Analysis Module, by statistical analysis and measurement, we supposed to gain the parameters of corrosion characteristic, such as quantity,area,size, etc, then evaluate corrosion feature according to Evaluation Criteria.The image analysis system of the surface of corrosion characteristic achieve the feature extraction, statistical analysis and evaluation of corrosion feature. Compared with manual observation and analysis, it is more efficient, accurate, fast and convenient. It also reduce the impact of human factors, so that the evaluation is more reliable and suitable for quantitative treatment, which can greatly improve the speed and accuracy in the assessment of surface corrosion characteristic, and it runs well in the application.