Imaging of Target Objects in Air and Underwater Environments by Optimization Algorithms

• Main Authors: Chung-kuan Huang, 黃種寬
• Other Authors: Kun-chou Lee
• Format: Others
• Language: zh-TW
• Published: 2008
• Online Access: http://ndltd.ncl.edu.tw/handle/88185394267721064883

碩士 === 國立成功大學 === 系統及船舶機電工程學系碩博士班 === 96 === The image reconstruction techniques have been applied to many areas, including medicine, computer, industry, traffic and human life, especially in the safety control. The image reconstruction also plays an important role in national defense. It will be well developed in the future. Therefore, this research was focused on the image reconstruction of both terrestrial and underwater environments. In the first part, the target is illuminated by plane electromagnetic waves. The scattered waves are then collected for imaging by signal processing techniques. This technology can be applied to the aircraft and ship detection. There have been many studies involved in the terrestrial microwave imaging. Through the imaging techniques, the target objects do not need to appear near the instrument. The microwave has been widely applied to non-destructive detection in industry, radar images in military, and image processing of human body in medicine. However, the microwave is not suitable in underwater environments due to its limited propagation distance. Alternatively, the acoustic wave can propagate very far and thus becomes a good candidate of underwater imaging. In the second part, we extend the terrestrial microwave imaging technique to treat underwater acoustic imaging. In this thesis, we utilize the moment method to calculate the scattered fields of objects. This method has been proved to be accurate and efficient by many researchers. The inverse scattering problem was successfully transformed into the minimization of a cost function. This cost function was minimized by the PSO (particle swarm optimization) method. As the minimum of the cost function is reached, the shape of the target object was reconstructed simultaneously. The PSO algorithm is an ideal searching principle on parameter space by using random choice. During the calculation, there is no differential operation. Nearly global optimum can be ficiently achieved by the PSO algorithm. In this study, the PSO was successfully applied to both the terrestrial microwave and underwater acoustic imaging.