| In recent years,the development of visible light communication(VLC)technology has attracted extensive attention.Optical camera communication(OCC)is a VLC technology that receives optical signals through image sensor(IS).Compared with wireless communication based on radio frequency(RF),OCC technology has the advantages of no RF interference,no spectrum authentication,no electromagnetic radiation and so on.Therefore,OCC technology has great application potential in complex environments such as intelligent transportation system,indoor and outdoor communication and positioning.At present,the research on OCC technology is mainly divided into coding and modulation,positioning and tracking,decoding and demodulation and so on.There are still some deficiencies in the research of decoding and demodulation:the existing research analyzes the performance of decoding algorithm through simulation or in an ideal environment without noise,without considering the influence of interference light source;The existing studies identify the state of LED by comparing the intensity of gray value with the binarization threshold,and do not consider the impact of the change of ambient light intensity on the overall gray value of the image.Based on the above problems,considering the environment of complex factors such as interference light source and light intensity change,this thesis makes an in-depth study on the decoding algorithm of OCC system.Firstly,this thesis analyzes the influence of ambient light variation and interference light source on the performance of decoding algorithm based on directional projection method;Then the algorithm is optimized by adaptive threshold and secondary decoding of"0"state LED;Finally,aiming at the problem that the optimization algorithm can not deal with the interference light source and has high dependence on the sensitivity of binarization threshold,a decoding algorithm based on multi features is proposed.The main work and contributions of this thesis are as follows:(1)OCC system optimizationBy reducing the number of positioning LEDs,the number of transmitting LEDs is increased from 12 to 14,and the effective data transmitted in each data frame is increased from 12 bits to 14 bits.By using a camera with better performance,the communication rate of the system is increased from 126.32bps to 789.47bps,and the communication distance is increased from 8m to 80m.(2)Analysis and optimization of decoding algorithm based on directional projection methodThe decoding algorithm based on directional projection method has the advantages of low overhead and high transmission rate.Firstly,this thesis analyzes the influence of the change of ambient light intensity on the performance of the decoding algorithm based on directional projection method in the environment of non-interference light source.The experimental results show that the algorithm recognizes the LED state through the binary image,and the sensitivity of the binary threshold is an important factor affecting the performance of the algorithm.Then,the decoding algorithm based on directional projection method is optimized from two aspects:one is to improve the sensitivity of binarization threshold through adaptive threshold,and the other is to reduce the dependence of the algorithm on binarization threshold sensitivity through the secondary decoding of"0"state LED.Experimental results show that the bit error rate of the optimized algorithm is reduced from 1×10-2to 2×10-3.(3)Research on decoding algorithm based on multi featuresAiming at the problem that the optimized decoding algorithm based on directional projection method can not deal with the interference light source,a geometric feature-based LED array detection algorithm is proposed.The interference light source is excluded according to the geometric features such as the width and aspect ratio of LED array,and the contour information of LED array is determined.In view of the low robustness of the optimized decoding algorithm based on directional projection method in the changing environment of ambient light intensity,the LED bit detection algorithm based on gray feature is optimized.The LED state is recognized by calculating the average gray ratio and gradient vector product in the LED area.The LED array detection algorithm based on geometric features and LED bit detection algorithm based on gray features are combined to form a decoding algorithm based on multi features.The experimental results show that the error rate of the algorithm in the complex environment is 5×10-4. |
PDF Full Text Request