Research Of Laser Insecticidal System Based On FPGA

In this thesis, based on FPGA platform, object detection and object tracking algorithm is implemented to process images in high speed and locate and track moving objects simultaneously, which has been focused in the field of computer vision. With the development of computer science and information technology, computer vision and pattern recognition algorithms are widely implemented in fields, such as industries, medical, military and agriculture. Computer vision is widely applied for its high speed, simultaneity and accuracy. And it is featured by the validity of object recognition. The core process of object recognition is divided into capture of image frames and image processing. The complicity of the two processes determines the efficiency of whole system. In this thesis, different object recognition algorithms have been analyzed, and a new image processing solution which is more adaptive to complex environment has been designed.On the designing guides, firstly, modular design has been implemented in this work to simplify the system and it is convenient to debug and adjust. Secondly, program with verilog language has been used to enhance the flexibility and accuracy of design. What’s more, FPGA is featured by its high integration and flexibility. By employing FPGA on our platform, the performance of the system has been enhanced and cost of the design has been reduced to minimum degree.On the processing of research in this thesis, firstly, a clip of video stream has been captured by CMOS camera, and each frame has been extracted by MATLAB. Then, ModelSim has been employed to realize several prevailing image processing algorithms, and the simulation is given. Next, MATLAB program has been employed to process the result data of the simulation. Finally, this thesis compared the results, chose the algorithm which leads to the best output, and improved it.A new improved highly efficient object recognition algorithm is proposed. It can be realized on FPGA platform. The whole system can be divided into six modules:image frame capturing module, image pre-processing module, image storage module, image reading module, image post-processing module and object extraction module. Through strict disposal, moving objects have been captured and extracted from the background environment, then the position information of the objects can be calculated, and make the laser to beat or track. The results of the experiments show that all the modules function can perform well, and the hardware perform with a high level of approximates, meeting the expected target.