摘要
为了给昆虫行为学的研究者提供可靠的统计资料,提出了一种从视频流中提取运动目标物的坐标值的方法,并设计实现了该算法。视频文件经过预处理后,提取静止背景,再确定每张图像中试验昆虫的区域,随后选取恰当的阈值,经过维纳滤波和中值滤波,从中提取运动目标物的特征。最后算出运动目标物在图像中的坐标值。通过对提取效果和统计结果进行分析,证明该方法有较高的提取率和可靠性,是实现视频资料数据化的有效手段。
To provide reliable statistical information for insect ethology researchers,a method to extract the coordinate values of a moving object in video files was proposed and the algorithm for it was designed and implemented in this study.The video files were preprocessed to extract static background and then get the region of insects in each image.After the appropriate threshold was selected,the images passed through wiener filtering and median filtering to extract the features of the moving objects.Finally,the coordinate values of a moving object in the image were calculated.Analysis on the result of extraction and statistical results proved that the method had a high extraction efficiency and reliability.Thus the proposed method is an effective method to transfer video files to data.
出处
《现代农业科技》
2010年第21期177-179,182,共4页
Modern Agricultural Science and Technology
