Research On Demodulation Algorithm In Optical Camera Communications

Optical camera communications(OCC)has gradually developed into one of the most promising wireless optical communication technologies due to its high penetration rate of communication equipment and its natural integration with emerging technologies such as the Internet of Things and Internet of Vehicles.The development will expand the application market of visible light communications(VLC)and enhance the commercial value of VLC.In addition,IEEE 802.15.7r1,as the official standard of OCC,has further promoted the development of OCC standardization.In the future,OCC is expected to be a candidate technology for VLC used in some low-speed communication services.Currently,most of the receivers used in OCC communication systems are cameras with lenses,and the optical components built into the cameras greatly increase the thickness of mobile devices,thus limiting the development of new ultra-thin cameras.In addition,a large number of scholars have conducted in-depth research on the use of lensless imagers for photography.Inspired by this,this paper proposes a novel lensless imaging communication system model.At the same time,considering that the current OCC related research is carried out in the static communication scenario,that is,the sending end and the receiving end of the communication are required to be in a static state at the same time.The reason is that the existing demodulation method cannot support the mobility of the terminal.Therefore,this paper studies the demodulation algorithm at the receiving end in our proposed system model,and aims to propose a mobile robust demodulation method to promote the application of our system in mobile communication scenarios.In addition,the OCC transmitter can be an LCD(Liquid Crystal Display)display,a digital signage or a projector,etc.in addition to LEDs.Information is embedded in the TV,monitor,billboard and even the screen of the projector in an invisible watermark In order to achieve data communication without affecting the video playback,this system is called invisible watermark screen-camera communication.Most of the existing demodulation schemes in the camera communication system with invisiblewatermark screens adopt block embedding in the watermark embedding stage,and then recover the data after region of interest(ROI)extraction and block processing at the receiver.Due to the change of shooting angle,shooting distance and other factors,the captured image will be deformed and the resolution will change,which will lead to more complex image processing,further affect the ROI extraction and the correct execution of the block,and ultimately lead to the error of data recovery.Therefore,the key of watermark communication system is to choose the appropriate watermark embedding method and design an effective watermark decoding algorithm.Aiming at these two problems,this paper proposes a new invisible watermark screen-camera communication decoding scheme.The main research contents of this paper are as follows:(1)This paper proposes a lensless imaging communication system model that uses a lensless imager as the receiver,whereas the optical components in the lensless camera receiver used in current optical camera communication(OCC)systems usually limit the overall thickness of the device.In addition,because the images captured by the lensless imager used in our system are blurred,which is also our advantage,we don't need a particularly clear image to locate the light source on the image,and we can implement decoding without complicated image processing algorithms.Considering that the decoding performance of ordinary image decoding algorithms varies greatly in a time-varying environment.To solve this problem,we propose an image decoding algorithm based on back propagation(BP)neural network.This algorithm can resist external interference(ambient light,bad weather),and it is also mobile robust.In order to make our decoding algorithm adaptive to the time-varying environment,we also designed an adaptive training sequence adjustment mechanism.Simulation results show that the image decoding algorithm proposed in this paper can provide our system with a good bit error rate(BER)performance.(2)This paper proposes a new invisible watermark screen-camera communication decoding scheme.This solution embeds a watermark globally at the sender side instead of block embedding.This avoids ROI extraction and block processing at the receiver,simplifies the complexity of image processing algorithms,and improves the effectiveness of the extracted watermark.After the watermark is extracted,we use the decoding algorithm based on the convolutional neural network to recover the data information.Convolutional neural networks have powerful feature extraction capabilities and adaptability.In a time-varying environment,the convolutional neural network can identify the subtle differences between various types of watermarks wehave extracted to ensure the accuracy of our data recovery.This paper also builds an invisible watermark screen-camera communication experimental platform to evaluate the bit error rate performance of our algorithm.This paper studies two kinds of communication demodulation algorithms of OCC system.The research results can provide new ideas for further research and future application of OCC.