机车走行部故障图像识别系统的研究

Research on Locomotive Running Gear Fault Image Recognition System

机车走行部的故障目前仍以人工检测为主,既有的车辆故障图像识别系统TFDS不能适应机车要求,亟待研发机车专用的走行部故障图像识别系统。根据机车走行部故障特征,提出了机车走行部故障图像识别系统构成及其原理,探讨了高清晰度图像的获取、低照度条件下图像的获取以及图像识别方法,并深入分析了建立机车走行部故障特征库,以及机车走行部故障评判及报警分级处理等问题。

Currently,the faults of locomotive running gear are mainly subject to artificial detection.The existing vehicle fault image recognition system TFDS can not meet the requirements of locomotives,and a running gear fault image recognition system specially used for locomotives is urgently needed.This article,based upon the signatures of the locomotive running gear faults,puts forward the makeup and principle of locomotive running gear fault image recognition system,discusses the acquisition of high resolution image,image acquisition in low illumination as well as image recognition methods.Meanwhile,it makes an in-depth analysis over the establishment of locomotive running gear fault signature library,and locomotive running gear fault judgment and evaluation as well as stage alarm treatment etc.