Evaluation for the Velocity of Debris Flow through Image Processing Techniques:A Case Study on Shen- Mu Monitoring Station

• Main Authors: Hsiu-Yi Ko, 柯秀宜
• Other Authors: Po-Whei Huang
• Format: Others
• Language: zh-TW
• Published: 2006
• Online Access: http://ndltd.ncl.edu.tw/handle/57755631564060865239

碩士 === 國立中興大學 === 資訊科學系所 === 94 === The landforms and geologic structure in Taiwan are peculiar, and terrains are easily broken and unstable. Washout and rainstorms occur very frequently during the typhoon season and result in debris flows and landslides in mountain areas. In recent years, the Soil and Water Conservation Bureau has devoted much effort on monitoring and investigation for the debris flows. Several monitoring stations and equipments were set up to observe, receive and transfer data to disaster management center. The real-time image data grasped by CCD camera is then transferred back to the SWCB via satellite communication network. It allows relevant personnel to grasp the real-time spot situation of every monitoring station during the typhoon season. Event happened in Shen-Mu Station, Nantou county, on July, second, 2004 is analyzed in this thesis. It demonstrated that human eyes can clearly observe the motion of debris flow, but the velocity must be calculated based on the flowing distance of water manaally. In order to strengthen the value of CCD camera images and applications of digital image processing, this research built a set of the rules to evaluate the velocity of debris flow. Technologies include feature extraction, image enhancement, spatial filtering and cell analysis are employed to build up the mechanism responsible for velocity interpretation. The result of velocity analysis can be taken as the reference and the evaluation of follow-up researches.