Research On Lucky Imaging Technology Based On FPGA

Lucky image technology is an post-processing method which is used to eliminate the influence of atmospheric turbulence in astronomical images and to obtain reconstructed high-resolution images.Although this technique is a simple and feasible method for image restoration,the shortcoming is obvious because the image restoration is performed within a certain period of time after the observation is completed.That is to say,the astronomical observers have little understanding of the real-time information of the captured images,Identify and correct possible deviations or mistakes in the observation in time.The solution to this problem is the real-time imaging algorithm lucky,but the traditional CPU based on the lucky imaging algorithm is difficult to real-time.However,in recent years,thanks to the rapid development of Field-Programmable Gate Array(FPGA)technology and the advantages of its parallel processing,a feasible approach for real-time or real-time luck imaging algorithms is provided.This dissertation mainly studies on this issue.In order to realize the lucky imaging algorithm in FPGA,the basic principle of lucky imaging algorithm is analyzed first.On the basis of not changing the basic principle of the algorithm,in order to meet the limitation of the logic resource of FPGA chip,Superposition of the modular design of each sub-algorithm for analysis and verification,and gives the project for the lucky imaging algorithm FPGA implementation of the specific programs and implementation process.In this system,the core part is composed of data reading module of Micro SD card,data reading and writing module of DDR3,lucky imaging algorithm module and image display module.Which read from the SD card image data,all the way to the DDR3 data read and write module processing,and then sent to the external DDR3 chip storage,and the other is input to the lucky imaging algorithm selection module for maximum gray value of the maximum gray value and maximum solution And then sort the value of the image into the registration module to obtain the first address of the desired image and send it to the DDR3 read-write data module,so that from the input of the image sequencenumber and the maximum gray-FPGA external DDR3 chip to read the image pixel required for the calculation of the overlay module to get the final reconstruction of high-resolution images;Finally,after the transformation of the gray transform module by the VGA driver module processing eventually displayed on the display.To test the feasibility of the proposed scheme and design method,a small-scale FPGA development board was used to build a lucky imaging system.The experiment was carried out by using the measured short-exposure binocular images.The results show that the high-resolution image obtained by this system is exactly the same as that based on the traditional CPU algorithm,but the fortune imaging algorithm can process more than 20 times faster than the traditional CPU.The implementation of this algorithm on FPGA provides an effective way for real-time or near-real-time fortune imaging technology.