视频监控中的考生异常行为识别

Exam Taker Behavior Recognition in Video Surveillance

为了提高考生考试行为识别的准确率,提出了一种视频监控中的考生异常行为识别方法ICanny-ABC-SVM。该算法从视频监控中提取考生行为图像,采用改进Canny算子对图像进行边缘检测;通过提取图像的不变矩特征,并将特征向量输入人工蜂群优化支持向量机中进行学习,构建考生行为分类器;运用仿真实验测试方法的性能。测试结果表明,此方法获得了较高的考生行为识别准确率与较快的识别速度,是一个性能较优的智能视频监控考生行为识别方法。

In order to improve identification rate of the exam taker behavior,an exam taker behavior automatic recognition method in intelligent video surveillance called ICanny-ABC-SVM is proposed in this paper.In ICanny-ABC-SVM,the image of exam taker is collected in intelligent video surveillance system,and canny operator is used to detect the edge of exam taker behavior image.Then,the invariant features of images are extracted and feature vectors are input into support vector machine to learn which parameters are optimized by the artificial bee colony algorithm.The performance is tested by simulation experiment.The experimental results show that ICanny-ABC-SVM has obtained higher identification rate and fast speed,and is an effective identification method in intelligent video surveillance system.