| Summary: | 碩士 === 國立臺北科技大學 === 電子工程系研究所 === 105 === Due to advances in technology, the development of autonomous car is growing fast. The application of 3D model technology also becomes more and more popular. It is very important to perform 3D reconstruction of environment according to 3D model because it can help autonomous car to detect the factor of surrounding environment by itself. Through the establishment of surrounding environment model, autonomous car has better ability to detect the surrounding environment. In this paper, we perform point cloud data process for 3D model reconstruction and the system detects the environment according to point cloud information. Afterward, ground filtering, RANSAC algorithm, region-growing algorithm, and point cloud analysis is performed to provide well model. The system could perform simple building model reconstruction. Besides, we also designed the system hardware architecture via the Cell-Based IC Design Flow. We implement the LiDAR data packet process part in a chip and optimized the hardware architecture. By speeding up the system algorithm, the chip is performed to achieve the 3D indoor scene reconstruction.
|
|---|
