Edge Detection and Geometric Feature Extraction of A Burner’s Jet in FNBZ Images Using Seeded Region Growing Algorithm
The method based jet trajectory is routinely used in a cold-state test of FNBZ (flow field near burner zone). However, the traditional measurement can not effectively obtain useful information from the jet images. In this paper, the approach of computer-aided diagnosis based image processing is attempted. The noisy nature of the images, as well as varying intensity, edge feature and centerline location of the jet under different boiler loads, defeat most automatic detection algorithms, making computer-aided diagnosis difficult. This paper applies the seeded region growing algorithm to jet images with manualed seed selection and realizes semi-automated detection of edge feature and centerline location of the jet in the FNBZ images, producing good performance in tests. With the high success rate of over 90％, refinement of the scheme could be applicable in cold-state tests of FNBZ as a tool in aiding diagnosis, with possible applications in analysis of flame images for combusting condition.
Flow field Image processing Burner Image segmentation Feature extraction Seeded Region Growing
Hui-yong Li Zhen Zhou Dong-lin Chen
School of Energy and Thermal Power Engineering, Changsha University of Science and Technology,Changs School of Energy and Thermal Power Engineering, Changsha University of Science and Technology, Chang
国际会议
2007杭州国际动力工程会议(The International Conference on Power Engineering 2007)
杭州
英文
2007-10-23（万方平台首次上网日期，不代表论文的发表时间）