火灾视频图像的边缘检测

针对火灾视频图像存在较多噪声的特点，介绍了图像边缘的概念和几种传统的边缘检测算子，对各种边缘检测算子的优缺点进行了分析，给出了一种图像阈值分割与边缘检测相结合的方法.该方法先对火焰灰度图像进行分割，采用阈值迭代算法找到分割的阈值，使火焰与背景分离，然后再应用传统的边缘检测算子.通过利用拍摄的火焰视频图像对传统的算法和给出的算法进行多次试验及图像的对比分析，结果表明，提出的边缘检测方法检测出的火焰边缘在完整性和清晰度上具有更好的效果.

Edge detection of fire video images

The concept of image edge and several conventional edge detection algorithms were introduced,and the advantages and disadvantages of relevant algorithms were analyzed.Aiming at the feature of much noises existing on fire video images,a method combining image threshold segmentation with edge detection was proposed.In the approach,the fire gray image was firstly segmented so as to detach the flame and the background,and the segmentation threshold was obtained using threshold iteration algorithm.And then,the conventional edge detection algorithm was used for flame edge detection.Using the flame video images,the repeated experiments and the comparison of the images were conducted for the given method and conventional algorithm.The result shows that the proposed method has better effect in integrality and definition for detected image edge.