Pole Piece Of The Battery Based On The Image Edge Extraction Defect Detection Research

Defects of electrode are the essencial factors to reduce the quality of battery, It is very important for Electrode defects inspection to improve quality of battery. Traditional Electrode defects inspection done by personal inspectors is out of satisfaction for user. Developing automatic surface defects inspection systems is a emerging demand of battery production enterprise.Edges of an image reflect the big parts of information of the image.They contain the main characteristics of image. Edge detection is a vital part of many image processing and pattern recognition systems. Typical areas of applications are image segmentation, stereo vision and identification of objects as in automatic target recognition. It is one of the most important parts in image processing. In this thesis, I proposed an Electrode defects edge detection method based on edge extraction. The details are as follows:1. According to the property of low contrast and small target for Electrode image, we proposed an edge detection method based on D-SUSAN algorithm, which is associated with the median filtering operation. The SUSAN algorithm has good anti-noise property, which can recognize the image edge very well. The case study indicate that this method is effective for my research.2. Due to extremely high pixel accuracy need for some Electrode defects inspection, The Canny operator is used to extract the defect edge, and its principle and implementation are discussed. The property of low contrast and small target for Electrode image is considered. A new Electrode defects edge detection method is proposed in this paper to adaptively determine the high and low thresholds of F-Canny operator using fuzzy maximum entropy by the edge gradient feature of the electrodes image.3. A filter method based on Mathematical Morphology is used for image denoising after edge segmentation.Some characteristic parameters are detected. Classify electrode defect by computing rotundity degree and slightness degree as characteristic vector. A solution plan for Electrode defects inspection is put forward. 4. Microsoft Visual C++.NET program is applied to develop this detection system. The modularization program method is used in the process of electrode image acquisition, image segementation, feature extraction and flaws detection.In order to validate the above inspection methods, we have detected the electrode image with some defects, which obtained from the factory. The detected results are definitely accurate. It is proved that this algorithm is adaptable for detection of battery quality.