【摘要】 针对玻璃绝缘子的多环状凹槽结构在图像中形成的光折射环带掩盖气泡、裂缝等缺陷,从而降低成像清晰度及易导致缺陷的误判、漏判的问题,提出通过频域滤波实现玻璃绝缘子图像增强的技术方案.分析气泡和裂缝的缺陷特性,对气泡采用同态滤波,对裂缝采用高频加强滤波,以达到抑制光折射环带所形成的复杂背景和突显上述两类缺陷的目的.试验结果证明,频域滤波能有效增强图像中缺陷部分的细节信息,利于后续的图像分割和缺陷识别,提高玻璃绝缘子质量检测的正确率.更多还原

【Abstract】 The structure of glass insulator is complex.Its circular grooves lead to refraction and reflection,which makes the image contain a lot of circular shadows.When the glass insulator was detected,these shadows cover bubbles and cracks and reduce the image definition.These shadows decreased the accuracy in quality detection.Aiming at this problem,the method of frequency filtering was presented.The defect characteristics of bubbles and cracks were analyzed and homomorphic filtering and high frequency emphasis filtering are adopted respectively to restrain the background caused by circular shadows and to enhance the defect.The test results indicate that the method of frequency filtering can enhance the details of defect effectively.It helps to image segmentation subsequently and defect recognition and improves the accuracy in the quality detection of glass insulators.更多还原