Application Of ASIFT Algorithm In Substation Equipment Image Recognition

Application of ASIFT algorithm in substation equipment image recognition is studied in this paper. First, theoretical analysis and comparison of the ASIFT and classical SIFT algorithm are made; Second, ASIFT algorithm was used in extracting the feature points of typical substation equipment image, and image feature matching under rotation, scaling, and noise was realized, providing help for automatic inspection, on-line monitoring of substation equipment; In addition, substation equipment tilt angle parameter inversion was researched using matching point result;At last, some common infrared image and visible light image of substation equipment are accurately matched, which provides more abundant and clear information for identification and assessment of the substation equipment. The specific work includes the following several parts:1. The classic SIFT algorithm in the field of local feature points extraction is studied. Operation speed and robustness in the changes of view were analyzed through programming simulation experiment.2. Performance differences of SIFT and ASIFT algorithm in substation equipment image recognition are analyzed.3. Application of ASIFT algorithm in substation equipment image recognition are studied, and both advantages and drawbacks are researched. Typical electric power equipment standard image was matched with images acquired in the inspection.ASIFT algorithm matching results were used to identify condition of electric power equipment such tilt angle, which can be applied in On-line monitoring.4. Applying of ASIFT algorithm in visible light image and infrared image registration was studied, and ASIFT registration method based on edge detection is also proposed in this paper.