Research And Implementation Of Optical Cable Protective Jacket Detection Technology In The Scene Of Three-wire Crossover Based On Image

Optical cable is a kind of telecommunications line.When the optical cable is erected,if the erected optical cable crosses or hangs on the wires that have been erected,the scene where the telecommunications line crosses or hangs on the power line is called three-wire crossover.In this scenario,even if the two are not in direct contact,the power line and the optical cable may occur coupling induction,which affects the life of the optical cable,and may also cause incoming power disturbance.Intrusion poses a threat to personal safety.Due to the above reasons,the part of the optical cable where the crossover occurs must be equipped with a protective cover.At present,the inspection of optical cable erection specifications is mainly done manually,resulting in a heavy workload for inspectors.The automatic detection of the installation situation of the optical cable protective cover from the images collected by manual construction is of great significance for the intelligent detection of the optical cable erection specifications in the three-wire crossover scenario.First,the research status of various directions in the field of power line detection is studied.After analyzing the characteristics of the actual images that need to be processed,the research status of related technologies in the field of sky segmentation is investigated.Based on the analysis of the process of manual inspection of the optical cable protective sleeve,the overall technical scheme of whether the optical cable protective sleeve is installed in the automatic detection image is determined.On the basis of ensuring low cost,a semi-automatic detection technical scheme was proposed and realized.In view of the characteristics of the collected image data,the sky segmentation method based on the energy function is improved,and other pixel areas that may affect the subsequent Hough transform calibration wire coordinates are removed.Then through the Hough transform,combined with the hierarchical clustering method,the complete coordinates of the various wires in the image are detected.Finally,using the feature that the wires and optical cables in the image dominate the number of pixels in the local area,the detection frame technology is used to complete the detection of the grayscale characteristics,color characteristics and pixel widths of the optical cables and wires,and then detect the optical cable protective sleeve in the image.The overall technical solution proposed and applied basically meets the needs of the Sichuan Telecom Company for detecting optical cable protective sleeves.After experimental testing,the accuracy rate of the overall technical solution can reach 71%,and the false positive rate is guaranteed to be 0%.After the end user's trial,the accuracy rate in actual detection reaches 67%,and the false positive rate is 0%.The results show that the detection effect basically meets the needs of users.