Railway Locomotive Image Positioning Algorithm

At present,China’s railway industry is developing rapidly,and the number of railway locomotives is increasing.Good operation and maintenance of railway locomotives is the basic guarantee for high-speed and stable development of railway industry.Locomotive repair and maintenance has also become an important guarantee for the normal operation of railway trains,at present,there are still a lot of traditional manual inspection work at locomotive maintenance points.There is a low efficiency and manual labor cost in manual detection.In the rapid development of the railway industry,the traditional mode of railway work urgently needs to realize automation and intelligent management.Locomotive automatic information management needs of this algorithm for the railway locomotive maintenance point,is an important part of locomotive identification system,mainly studies the digital image processing and machine learning techniques in railway locomotive localization algorithm based on image region.It is an important algorithm module of the railway locomotive inspection system(TLDS)of the locomotive overhaul point of Wuhan Railway Bureau.The algorithm mainly uses digital image processing technology and machine learning technology,precise positioning number area for railway locomotive of various types of samples in the picture,at the same time to railway locomotive regional image processing,and the number of regional positioning pictures as a result output,for subsequent identification processing.According to the actual demand,we design and implement a multi algorithm matching algorithm for railway vehicle number area image location.The main work is as follows:(1)According to the characteristics of railway locomotive image,design a preprocessing step,the adaptive enhancement algorithm scales and ROI area image selection,image(logarithmic transform enhancement,Retinex enhancement algorithm,Gamma correction,noise processing(Gauss),fuzzy median filtering,filtering,image guided)dimensionality reduction and feature extraction of color area work,the original railway locomotive sample image is preprocessed to reduce the noise of image information,reduce the difficulty of image feature extraction,to ensure the effective operation of the subsequent location algorithm,and can effectively reduce the system resources occupancy rate,improve the efficiency and robustness of the algorithm.(2)The number of coarse positioning algorithm design,to find the candidate region number.By using the algorithm of color feature extraction,image edge information detection(Sobel operator,Canny operator,Sobel operator in HSV channel)and maximum stable region algorithm,we can effectively extract the features of the image.At the same time using the close operation method of digital image processing the extracted feature integration,image feature information through contour search will integrate the good effective judgment to determine the candidate region number area.(3)The number of accurate positioning algorithm design,Tilt correction candidate region of the adaptive to a certain degree in accordance with prior knowledge on certain conditions,then the scanning method to judge and support vector machine to determine 2 kinds of ideas for the candidate region judgment,selected the correct number of regional,by train after the last time to determine the area of precise positioning sure,the final number of regional,as the whole algorithm results are output,complete the whole process of algorithm.The algorithm is developed using Microsoft Visual Studio 2010 and computer vision library(OpenCV2.4.9),and is packaged as a form of dynamic link library(DLL).Through interface calls,it ensures the confidentiality of algorithm and the convenience of improvement and optimization.Through a large number of online testing and field testing,and gradually improve the function and stability of the algorithm,realizes the railway locomotive regional precise positioning function,through the batch test,the positioning accuracy can be up to 97.6%,reached the practical requirements.