Study on Fringe Pattern Processing and Recognition of Photoelastic Stressmeter Based on Wavelet Transformation and Neural Network
The fringe images of photoelastic stressmeter can only be observed and recorded artificially in field applications at present. In order to eliminate ill effects of artificial factors in results observed, a new pattern photoelastic apparatus which can film fringe images is developed. The fringe image filmed are decomposed, reconfigured, segmented, compressed, and denoised through wavelet transformation, and after these processing mentioned above they are recognized by a 4-layer BP neural network recognition system, which is established beforehand. Recognition results of fringe image filmed indicate that the processing and recognition system established has strong practicability and good recognition results, and can be applied in project sites.
wavelet transformation photoelastic stressmeter fringe image processing image recognition neural network
JIANG Weidong ZHAO Kui ZHI Xueyi
Business School of Central South University, Changsha, 410083, China School of Environmental and Architecture,JiangXi University of Science and Technology,Ganzhou,341000
国际会议
The 2007 International Symposium on Safety Science and Technology(2007采矿科学与安全技术国际学术会议)
河南焦作
英文
2358-2362
2007-04-17（万方平台首次上网日期，不代表论文的发表时间）