Design And Research Of Pointing System Of Rotating Double Prism Based On FPGA

The rotating double prism system is a promising choice for portable and mobile systems due to its compact size,low power consumption,large field of view and excellent dynamic performance.The rotating dual prism pointing control technology changes the direction of the beam by independently rotating the two prisms,and can change the imaging axis and the direction of the outgoing beam.It has been successfully applied in technical fields such as embedded vision,UAV search and rescue,measurement and positioning.For practical application of rotating double prism scanning and pointing,exploring the relationship between the rotation angle of the two prisms and the spatial position of the outgoing beam is the key technology of the current double prism target pointing system.The traditional method relies on the angular position of the two prisms detected by the sensor,calculates the azimuth and pitch angle of the target space according to the vector refraction law,and then updates the prism rotation angle to make the target image at the center of the detector.However,this process is highly dependent on the parameters of the prism system and the two prisms are in ideal positions,so the error of the prism system cannot be ignored.In order to address the above problems,this paper first discusses the structure of the rotating double prism and designs an opto-mechanical model,combined with the double prism scanning model,and proposes a method of adjusting the visual axis using pixel error feedback.Combining this method and the opto-mechanical model,a software system is designed,which mainly consists of two parts: image detection and prism movement.According to the characteristics of the specific target and FPGA,combined with the difficulty of the hardware implementation of the algorithm,the image filtering algorithm,the sobel segmentation algorithm and the target center extraction algorithm are specially improved.In addition,the system mainly fits the target center from the detected target edge image,and determines the difference between it and the center of the field of view to obtain the pixel error.The pixel error of two adjacent frames are compared and the prism is driven by the stepping motor to independently rotate to the system.The sight axis moves to the center of the target.In order to verify the correctness of the system and the method,the corresponding hardware/components were selected and the rotating double prism experimental platform was constructed.The high-performance hardware and the sophisticated electromechanical structure are the practical basis to ensure the stable operation of the system.In this experiment,a prototype of the whole machine was designed with a reduction gear transmission mechanism and a concentric double prism assembly.A suitable prism element,a single image sensor chip and a stepping motor were selected as the actuators for the system’s beam deflection and image acquisition.At the same time,the main controller used the cyclone 4E chip,which not only has a wealth of internal logic gates,but also uses IP cores to implement complex algorithms,which can meet the requirements of the two key functions of image recognition and motion control.Algorithm implementation is the focus of this system.This article uses Verilog language to implement the corresponding algorithm on FPGA and gives an introduction to specific principles and operations.The function realization of the system is mainly the logic circuit design of the hardware drive circuit and the specific algorithm.According to the top-down module design method,the system function can be subdivided into different modules to realize the corresponding function or algorithm.The input/output interfaces of the modules are interconnected.Debug different functions according to different modules to further optimize the parameters of the system.In this article,the system engineering is divided into various modules such as pixel format conversion,filtering and noise reduction,Sobel edge detection,target center fitting,and motor drive.At the same time,the algorithm design diagram of each module and the related RTL port diagram are given.After the design is completed,this article uses onboard debugging.Each module is debugged first and Signal Tap Ⅱ is used to display its waveform.After debugging,the designed system and the method were verified and the experimental results were compared and analyzed.