| Edge detection is one of the most elementary techniques in the field of image processing and computer vision. How to detect the edge of the image accurately and quickly is the focus of these research fields all the times. Adaptive edge detection is characterized by changing the algorithmic threshold dynamically at runtime, which is of great flexibility and practicabilityNowadays, DSP system for image edge detection has been used to many fields. With the development of the digital system, the design and implementation of system need more flexibility and scalability. The flexibility and scalability of DSP is inferior to that of FPGA. The technology of active partial reconfiguration on FPGA can reconfigure the hardware itself with flexibility. So this thesis implements image adaptive edge detection by using active partial reconfiguration technology.In this thesis, the active partial reconfiguration, the reconfigurable computing system modeling and the operation of Impulse C which is used to implement hardware task quickly are delved deeply. In addition, an adaptive edge detection system based on active partial reconfiguration technology is designed and implemented. The system implements the following functions:image acquisition, image edge detection, threshold changing, dynamic partial reconfiguration and web services. After functional expansion, this system can be used in those requiring the image of on-the-spot requisition and accuracy, such as intelligent traffic management, automatic cleaning of drifting aquatic surface contamination, geometry automatic measurement during the manufacture and assembly of mechanic kits.At the end of this thesis, the system is proved to be capable of fast and flexible detection of image edge, through several tests, such as the hardware self-adaptive detection of image edge according to the photograph taken at the same place but under various light conditions, as well as the test of executing time of each task during the running process of the system. |
PDF Full Text Request