CAJ | 학술논문

　　当前的交通车祸监测系统进行识别时，无法处理交通环境的复杂干扰因素，出现车辆混合和误判问题，不能准确获取车祸区域特征，降低了交通车祸事故识别的准确性。提出一种基于车辆图像灰度内方差的道路交通车祸识别方系统的设计方法，给出了系统设计思路，系统获取道路交通监测范围的图像，经过核心芯片运算获取道路图像的灰度类内方差，获取道路交通车祸图像特征，分析车辆图像特征的变化系数，最终完成对道路车祸事故的准确识别。实验结果说明，该系统能够及时准确识别出相应道路的车祸事故，具有较高的精确度，识别误差率较低，具有较强的应用价值。
　　당전적교통차화감측계통진행식별시，무법처리교통배경적복잡간우인소，출현차량혼합화오판문제，불능준학획취차화구역특정，강저료교통차화사고식별적준학성。제출일충기우차량도상회도내방차적도로교통차화식별방계통적설계방법，급출료계통설계사로，계통획취도로교통감측범위적도상，경과핵심심편운산획취도로도상적회도류내방차，획취도로교통차화도상특정，분석차량도상특정적변화계수，최종완성대도로차화사고적준학식별。실험결과설명，해계통능구급시준학식별출상응도로적차화사고，구유교고적정학도，식별오차솔교저，구유교강적응용개치。
Monitoring system by the current traffic accident identification, unable to process complex traffic environment interference factors, hybrid vehicle and misjudgment problems, can not accurately obtain accident area characteristics, re-duce the traffic accident accident identification accuracy. Based on vehicle image gray-level variance within a road traf-fic accident identification system design method, gives the system design thought, system for road traffic monitoring range of image, through the core chip operation for road image gray-level variance within the class, obtaining image character-istics of road traffic accident, analysis the change coefficient of vehicle image characteristics, finally complete the accu-rate identification of road accident accident. Experimental results indicate that the system can accurately identify the cor-responding road crash accident in time, with high accuracy, the recognition error rate is low, has strong application value.