Abstract: In view of the difficulty in the identification of transformer bushing, a method of identification of transformer bushing condition based on image matching is proposed. Firstly, on the basis of Guided filter Retinex image enhancement algorithm is used to the acquisition of transformer bushing image preprocessing. Then, the improved SIFT feature matching algorithm is used to accurately locate the measured image. Based on the maximum entropy thresholding（OET） method, the edge of casing is extracted. Finally, the recognition device is designed, and the operating condition identification is carried out. Results and analysis are given. And the method is applied to the engineering field, which reflects higher practical value.

Key words: image matching, transformer bushing, Scale-invariant Feature Transform（SIFT）, condition identification

摘要： 针对变压器套管工况识别难度大的实际，提出了基于图像匹配的变压器套管工况识别方法。首先，采用基于Guided滤波的Retinex图像增强算法对现场采集的变压器套管图像进行了预处理，然后采用基于改进的SIFT特征匹配算法对待测图像进行精确定位，利用基于最大熵阈值法（OET）图像分割方法提取了套管的边缘。最后，设计了识别器，进行了运行工况识别，给出了结果和分析，并将该方法应用于工程领域，体现了较高的实用价值。

关键词: 图像匹配, 变压器套管, 尺度不变特征变换（SIFT）, 工况识别