摘要
为了降低粮食在仓储环节因多种原因产生的损耗,需要及时准确地获取粮仓内的粮面状态,以便实时对异常的状态进行处理.提出一种基于感知哈希与连通域分析算法的粮面异常状态识别方法,定时或实时地采集粮仓内的粮面图像,存储为视频流并进行对比.通过感知哈希算法将采集的粮面状态图像与原始状态图像进行比对得到相似度,将相似度与事先设定的阈值进行比对,来判断粮面状态是否存在异常.若存在异常时,通过连通域算法进行异常区域的识别计算,该识别方法能够及时发现粮仓内粮面是否存在异常,并且能够精准定位异常区域,能够及时预防粮食仓储环节的损耗.
Abstract
This paper proposes a method for identifying abnormal grain surface states based on perceptual hashing and connected domain analysis algorithms.By collecting images of grain surface in a grain warehouse at regular intervals or in real-time and storing video streams for comparison,the perceptual hashing algorithm compares the collected grain surface state images with the original state images to obtain similarity.The similarity is then compared with a pre-set threshold to determine whether there are any abnormalities in the grain surface state.If there are abnormalities,the connected domain algorithm is used to identify and calculate the abnormal areas.This recognition method can timely detect whether there are abnormalities in the grain surface in the grain warehouse and accurately locate the abnormal areas,which can prevent loss-es in the grain storage process in a timely manner.
维普
万方数据