Abstract
The catenaries must guarantee the constant electricity to the trains, so that the safety and the estimate of fatigue degree and the inspection of abrasion degree should be done rapidly. This thesis proposes the system that can manage the performance/failure of the catenaries using the image process as the solution for the weak points of the existing inspection such as the decrease of immediateness and the lack of constancy in the human resources. This study simulates the ALD using VHSV (Virture HSV) which is the unreal HSV images, not getting the real HSV in ages to the image processing technique that repeats the erosion and the expansion of the images as the methods that can monitor the slight/critical defects of the catenaries as dealing with the result of the catenaries inspection images. The final ALD (Abnormal Line Detection) system is designed based on this simulation. I have demonstrated it with the VHSV (Virture HSV) virtual images as the materials of the test and inspected it through testing the defects of the catenaries for the thesis.
