A Nested Texture Inspired Novel Image Pattern Based Optical Camera Communication

Abstract

Pattern recognition is a vital component of display-based optical camera communication. However, as the distance between the transmitter LED panel and the receiver camera increases, the visibility of the pattern reduces. Decoding each pattern requires extensive computational resources. This paper proposes a nested texture-inspired novel binary hierarchical image pattern classification-based modulation for optical camera communication to provide efficient pattern classification and achieve longer communication distance. Custom $8\times 8$ patterns are generated at the transmitter based on input data using nested outer and inner patterns. On the receiver side, multi-threaded pattern matching is performed simultaneously for inner and outer patterns for faster decoding. A simulator is built to test the performance of linear search and two variants of binary search for pattern matching. A hardware prototype and Matlab app are designed to perform experiments to test the performance of the proposed technique in a real-world environment. Experiments were conducted to select optimal camera parameter values for the best signal to noise ratio (SNR) and to analyze the impact of pattern intensity profile on decoding at various distances. The proposed pattern communication technique showcased 48.6% and 42.5% improvement in bit error ratio (BER) compared to Data Matrix and QR Code Pattern-based communication techniques, respectively.

Keywords

• Internet of Things
• optical camera communication
• parallel computation
• pattern communication
• texture classification