Design And Implementation Of UAV Line Patrol Image Management And Image Duplicate Checking System

Using Unmanned Ariel Vehicle to inspect the operating situation of power grid system has largely improved the efficiency of power grid maintaining works,but it also brings large amount of data,including videos,pointclouds,and images.How to manage those data effectively while being able to handle data duplicate,and handling system defects intelligently is a developing hot topic in software engineering and automatic engineering societies.In order to handle those situations,we have proposed a design of the system framework for duplicate reduction based on image segmentation and image search.The proposed system firstly separate the electrical transmission pole and electrical transmission lines from natural and civil backgrounds with the help of spatial information constrained C-means clustering algorithm,and then use the background content to identify image groups.After that the visual code book need to be constructed where visual code is generated from clustering of SIFT features extracted from images in each image group.To identify image duplicate caused by images from other image groups,we compare representation of the image in the visual code book from that of the mean representation of images in the same image group,while using Hoff voting based geometric transformation verification method to discriminate image duplicate caused by image transformation of images in the same image group.What’s more,in order to detect system defects,taking electrical insulator hammer erosion as an example,we firstly detect electrical insulator hammers using Histogram of Gradient features and Support Vector Machine,and use the ratio of the area of corrosion spot to area of electrical insulator hammer to determine how corroded the electrical insulator hammer is.The corrosion spot is indicated after thresholding the pixel intensities in the HSV space of the image and transform back to RGB space,provided the corrosion inspector has decided that it is a corroded one.All the proposed frameworks and algorithms have been validated with images taken from UAVs,which turns out that our proposed frameworks and algorithms are quite effective.Lastly,we have given a design of the software architecture of the whole image managing system,with detailed description of individual modules,including the image recording module,the image duplicate reduction module,the client managing module,the defect inspector and the reportgenerator.