| At present,WLAN microscope camera is playing an important role in interactive teaching,integrated circuit manufacturing and other fields.However,the current WLAN microscope camera has some problems,such as insufficient video resolution,not using lightweight system,high power consumption.This paper designs and implements an ultra-low power WLAN microscope camera with HI3518EV300 chip as the main control chip,STM32L0 chip as the standby control chip,and lightweight LiteOS system.Firstly,the paper studies the low power architecture design of WLAN microscope camera,which lays a theoretical foundation for the subsequent camera hardware and software design.In terms of hardware,this paper discusses the selection standards of three components that consume more energy in WLAN microscope camera;In terms of software,the selection of low power video processing algorithms in WLAN microscope camera is studied,and the power performance of three mainstream video coding algorithms is compared.In terms of overall power management,this paper discusses the dynamic power management of WLAN microscope camera,analyzes three state switching strategies and applicable scenarios,and calculates the optimal threshold time in the timeout switching strategy through mathematical modeling.In terms of hardware,the paper studies the selection of components of the WLAN microscope camera and the design of circuit schematic diagram,and completes the production of PCB.In the aspect of network,the paper designs a complete algorithm of camera and user terminal cooperation.Firstly,based on RTSP and Hisilicon’s video buffer,the paper completes the design of high-definition network video stream push.Then,this paper studies the parameter synchronization design of multi-user terminal and camera based on UDP.Finally,the paper studies the detection function of the camera by the user terminal based on IP broadcast.In the aspect of low power architecture design,the hierarchical state model of the WLAN microscope camera is designed by using the hierarchical finite state machine framework in the hisilicon’s development package,the five states of the camera and the switching logics among them are determined,and the state message processing mechanism is designed.The overall power consumption test results of the camera show that the proposed low-power architecture greatly reduces the power consumption of the camera.The power consumption of the camera in sleep state is 85%lower than that in working state,and the power consumption of the camera in working state is 38%lower than that of similar products.Paper also has carried on the software implementation and test on ISP function:in terms of the white balance,this paper proposes a ROI white balance algorithm based on the prior knowledge,tests a variety of microscope of different light sources,fits the color temperature curve,selects white points according to the relative position of the pixels with color temperature curve,calculates RGB gain according to the gray world algorithm based on the selected white pixel,completes the ROI white balance.The test results show that the algorithm can achieve accurate white balance for microscopes with different color temperature sources.In contrast enhancement,a contrast enhancement algorithm based on Gamma curve is proposed.In order to make up for the shortcomings of histogram equalization algorithm,the statistical information is used to correct the gray distribution function,and a new gray accumulation function is used to calculate the Gamma curve.The test results show that the algorithm smoothes the histogram effectively and achieves the effect of contrast enhancement. |
PDF Full Text Request