Abstract
II
Abstract
With the continuous enhancement of China's comprehensive national strength, the
rapid development of all walks of life, the continuous expansion of scale and strength,
followed by higher requirements for the reliability and security of our power grid, which
requires us to take appropriate measures and means to improve the reliability of the power
grid and the control ability of the power grid operation, when abnormal, staff can be timely
warned. To judge equipment defects or abnormal operating conditions of equipment and
take appropriate measures to eliminate defects, live detection profession plays the role of
"physical examiner" in power grid, and gradually becomes an important detection project
in power grid. Infrared detection technology, as one of the important detection means, has
been widely used in power system. The application of this technology greatly improves the
early warning ability of power system faults and defects, avoids corresponding accidents
and reduces unnecessary blackouts to a large extent. However, the current infrared
diagnosis technology still needs to rely on the experience and technical level of the
inspectors to analyze the infrared image in most cases. Under special circumstances, it is
difficult to guarantee the real-time and accuracy of fault analysis and judgment. In this
paper, image processing technology is used to realize the fault detection of substation
electrical equipment and improve the detection level of infrared diagnosis technology.
The main contents of this paper are as follows:
Firstly, the development history of infrared detection technology is briefly introduced,
and the principle of infrared imaging is introduced, and the imaging characteristics of
infrared image are analyzed based on this scientific basis. The common image denoising
methods, such as mean filter, adaptive filter and wavelet transform, are introduced, and
their denoising effects are analyzed and compared. It is determined that wavelet transform
and wavelet transform are selected in this paper. Mean filter combined with denoising
method. Secondly, aiming at the low contrast of infrared image, the method of image
enhancement is introduced, and the histogram equalization method is used for
experimental detection. Thirdly, in the technical means of image segmentation, the
methods of edge-based segmentation, threshold-based segmentation, watershed algorithm
and clustering algorithm are introduced in detail, and their respective algorithms are
introduced and experimented. Finally, a method based on image gray value is proposed to
万方数据
