Research On FPGA-based Visible Light Video Transmission And Positioning System

Visible light communication(VLC)is a branch of optical wireless communication(OWC).In a stricter sense,the core idea of VLC is to realize joint illumination and data transmission by modulating light sources.It has the advantages of green environmental protection,abundant spectrum resources,anti-electromagnetic interference,and high security.Therefore,in the context of increasingly scarce wireless spectrum resources,visible light communication technology has increasingly become one of the research hotspots in the field of wireless communication.Considering that in some special scenarios such as underground rescue,we require communication and position.So,a visible light-based video transmission system and an image sensor-based visible light positioning system are studied in this paper to meet these needs.The main work of this paper includes the following aspects:1.The working principle and characteristics of light emitting diodes(LED)and photodiode,commonly used modulation technology in visible light communication are introduced,through the channel model of indoor short-distance line-of-sight(LOS)link,the optical power distribution,signal-to-noise ratio and bit error rate,etc.Simulation analysis was carried out.2.A visible light video transmission system with a rate of 5Mbit/s is developed.It includes the collection,storage and encoding of video image data,the design and production of LED drive circuit and photoelectric conversion circuit.The video signal is collected by an analog-to-digital converter(AD),an adaptive filter based on the least mean square error(LMS)equalization algorithm is proposed to apply to the visible light video transmission system,and the filter module is implemented by a field programmable gate array(FPGA),which improves the system performance.In the demodulation module,the video signal is first recovered and extracted,and then the video signal is transmitted to the PC through the Ethernet output module based on User Datagram Protocol(UDP),and finally the real-time clear video display is realized on the PC.3.A visible light localization system based on image sensor is built,and a dual light source localization technology based on support vector machine(SVM)algorithm is proposed.At the sending end,the position information is generated by FPGA,encoded and loaded into the corresponding LED driving circuit,and the image sensor at the receiving end captures the images of the two LEDs.In actual positioning,due to the uneven distribution of optical power on the receiving plane,the received optical power changes when the positioning point moves,which affects the exposure and exposure time,and then makes it impossible to accurately extract the stripe features.In order to solve this problem,a stripe feature extraction method that selects column pixels is proposed.On this basis,the SVM learning algorithm is used to realize the classification and accurate identification of LEDs,and the three-dimensional imaging positioning algorithm of visible light is used to obtain the actual position of the target.Finally,the positioning accuracy of less than 2cm is achieved within the space range of 30cm×30cm×50cm.4.The influence of parameters such as transmission distance and communication rate in the video transmission system on the transmission performance is tested and analyzed,and the relationship between the bit error rate and the above parameters is drawn.The positioning error in the visible light positioning system is analyzed,and the performance test of the movement trajectory of the positioning target is carried out.