摘要
根据开放环境下火灾烟雾检测的实际应用要求,该文提出了基于小波能量的轮廓抖动性烟雾检测算法。该算法首先使用卡尔曼滤波检测烟雾运动区域,其次对运动区域进行二维小波变换求出前景的轮廓。最后,在动态检测中将轮廓的小波能量和轮廓的不规则性作为两个特征,输入到贝叶斯分类器,通过概率判定是否存在烟雾。此算法可提高烟雾检测的准确性,降低误检率,能较好的满足开放环境下火灾烟雾的监测预警。
According to the practical application and requirements of smoke detection in an opening environment,this paper presents the smoke detection algorithm based on the contour jitter of wavelet energy.The algorithm firstly detects moving area by using the Kalman filter,then gets the counter of foreground with the two-dimensional wavelet transform.In the dynamic detection,the wavelet energy and the irregularity of contour are viewed as two features,which are input to the Bayesian classifier for judging smoke whether exist or not through the probability.This algorithm can improve the accuracy of the smoke detection and reduce false detection rate,meet smoke monitoring and forecast in an opening environment.
出处
《杭州电子科技大学学报（自然科学版）》
2011年第5期183-186,共4页
Journal of Hangzhou Dianzi University：Natural Sciences
基金
国家自然科学基金资助项目(60775016)
浙江省重大科技专项资助项目(C13062)
