Design And Implementation Of 4K Microscope Camera Based On Hi3519A

With the increasing use of microscope cameras in the fields of life science,cultural relics,industrial manufacturing and so on,the demand for their performance has improved increasingly.At present,there is still a certain gap of developing microscope cameras between domestic and well-known foreign companies;therefore,based on the Hi3519A platform,this paper has developed a high-performance 4K microscope camera with its own intellectual property,which has great research valueIn terms of the hardware of the camera,this paper has completed the design of connection and the selection of main chip,RAM,Flash,image sensor,etc.In terms of the software of the camera,the overall design and functional development from kernel space to user space are completed.In terms of video stream processing of the camera,the algorithms for optimizing image quality have been analyzed from four dimensions of brightness,color,distinctness and definition,and corresponding optimization algorithms applicable to the 4K microscope camera are implementedThe video preview of the camera designed in this paper supports both HDMI and Gigabit Ethernet modes.And HDMI mode supports not only 4K resolution display but also 1080P resolution display,in which the camera can automatically identify the EDID of the HDMI display and switch to the best output resolution;Gigabit Ethernet mode not only supports the camera directly connected to the computer via a network cable,but also supports the camera and computer connected to the same router to realize the interaction of image data.The graphical user interface of HDMI mode of the camera in this paper is developed by Qt,in which feature-rich measurement,image comparison,picture-in-picture and other functions meet customers’ actual needs.The image encoding and decoding of the camera designed in this paper supports image encoding and decoding with the resolution of 3840×2160 as well as video encoding and decoding with 4K@60fps.When encoding the image,it supports saving the thumbnail and all measurements merged into the image;when decoding images,the decoding speed is optimized;when decoding the video,the frame-by-frame function is a special function developed by this design in combination with the requirements of the application of microscope cameras,which is often used to analyze transient changes of the object under the microscope.As for the automatic exposure module of the camera,this paper has optimized automatic exposure strategies of the camera,increasing the maximum exposure time from 0.033 seconds to 7 seconds,providing four automatic exposure strategies for users to choose flexibly,and using the gray histogram for objective analysis,which can automatically adjust the brightness of the image to the appropriate level accordinglyAs for the white balance module of the camera,this paper utilizes Bayer data to correct the gain values of the R,G,and B channels,by matching and verifying the data,having avoided the waste of time due to erroneous data and enhanced the robustness of the white balance algorithm.Combining the objective analysis of the average value of U and V components of the image with the statistics of the running time,the efficiency of the optimized algorithm in this paper is proved.As for the video denoise module of the camera,this paper proposes a 3D denoise algorithm based on exposure gain,after testing a large number of motionless and motional scenes and analyzing the corresponding PSNR values,proving that the optimized algorithm can achieve good denoise effect in both motionless and motional scenes.In this paper,the video of the camera has been edge-sharpened,by improving the traditional unsharp masking algorithm,adjusting the intensity of edge sharpening according to the ISO value,and through the objective evaluation of Tenengrad gradient function,demonstrating that the optimized algorithm is better than the traditional unsharp masking algorithm.