摘要
研究了架空线路瓷绝缘子破损视觉检测问题.基于瓷绝缘子局部区域特征,采用三层BP神经元网络从复杂背景中识别出目标绝缘子并定位,然后采用扫描分析绝缘子图像窗口纵向剖线模式的方法,判定绝缘子的完好状态,最后进行了仿真和实验.实验验证了绝缘子破损视觉检测方法的可行性.
Vision Inspection System for porcelain insulator fault is researched in the paper. First, based on its part region feature, the insulator being inspected is recognized and located against its background via a 3-layer BP neural network; then by scanning the cropped image window that contained the insulator and analyzing the mode of vertical section line, the status of the insulator is determined. Experiments .show that the proposed technique is practical.
出处
《高技术通讯》
CAS
CSCD
北大核心
2005年第8期35-38,共4页
Chinese High Technology Letters
基金
国家高技术研究发展计划(863计划)