摘要
火焰图像边缘检测是火焰图像检测系统研究的基础。将SUSAN算法引入到火焰边缘检测之中,并针对SUSAN算法中人为设定阈值在一些特殊场合下无法有效提取图像边缘的问题和运算量过大不适用于实时场景的缺点,通过引入目标区域判别和自适应阈值选取,提出一种改进SUSAN算法,解决上述两个缺陷并对该算法进行仿真。实验结果表明,该算法可以有效地提高火焰检测的准确率,排除干扰源,并具有良好的自适应性。
The flame image edge detection is the base to research the flame image detection system. The flame image edge detection model of an improved SUSAN algorithm is proposed in this paper. The SUSAN algorithm has some drawbacks： 1） the image edge cannot be effectively extracted in some special occasions due to the man-made threshold setting; 2） the too large computation quantity makes it not to be applied to the real-time occasions. Therefore, an improved SUSAN algorithm is proposed, by which the target area distinguishing and adaptive threshold selection are introduced to overcome the shortcomings of SUSAN algorithm. Experimental results show that the algorithm proposed in this paper can effectively improve the accuracy of flame image edge detection, eliminate the interference sources, and has perfect adaptability.
出处
《现代电子技术》
北大核心
2015年第5期58-61,共4页
Modern Electronics Technique
基金
教育部高等学校博士学科点专项科研基金(20126120110008)
教育厅提供专项科研项目(2013JK1144)
西安建筑科技大学校青年基金(QN1429)
