Full-color single pixel imaging based on spatial multiplexing Walsh-Hadamard tricolor coding strategy

•We propose a full-color single pixel imaging method based on spatial multiplexing tricolor coding strategy to realize imaging of color objects, which has only one single pixel detector and achieve high quality.•Compared with frequency division multiplexing, color image reconstruction can be achieved without complex reconstruction algorithms and complex optical path structures.•Pixel counts of the reconstructed image is consistent with the original image, with complete color information preservation and high imaging quality. Single pixel imaging (SPI) is an optical imaging technique that reconstructs target information based on active lighting and energy fluctuation principles. In this paper, a full-color single pixel imaging based on spatial multiplexing Walsh-Hadamard (SMWH) tricolor coding strategy is proposed. In design stage of structured light, utilizing spatial multiplexing technology, tricolor Walsh-Hadamard based structured light is encoded in a certain order and placed at different positions in the entire spatial domain without overlapping each other. A series of energy information obtained by a single pixel detector can be reorganized into a Fourier spectrum, which contains color information. In reconstruction stage of full-color target, the information of tricolor channels can be fully extracted and reconstructed into high-quality color images through a special algorithm that matches the encoding method. The proposed spatial multiplexing single pixel imaging is different from frequency division multiplexing method, which can provide a novel imaging scheme for obtaining full color images with energy information and color illumination. On this basis, we quantitatively analyzed impacts of different pixels and tricolor encoding methods on reconstruction quality by related evaluation indexes. Simulation and experimental results demonstrate the feasibility and effectiveness of the proposed full-color imaging method.