Dynamic occlusion removal in single-pixel imaging system based on self-check

Single-pixel imaging (SPI) system modulates the object with a series of patterns, records corresponding measurements bucket detector and forms an image by algorithm compressed sensing. In this process, if other objects enter into field view SPI, accuracy measurement will be seriously affected, quality reconstructed decrease. Owing to randomness reflectivity shape occlusion, it is difficult effectively separate disturbed part from signal. To solve problem, we propose self-check method based on characteristics Hadamard matrix, that is, using values verify correctness Usually when matrix as in divided complementary positive pattern negative pattern. The these two patterns are subtracted form (the difference value marked <inline-formula><tex-math id="M1">\begin{document}$ l $\end{document}</tex-math><alternatives><graphic xmlns:xlink="http://www.w3.org/1999/xlink" xlink:href="3-20221918_M1.jpg"/><graphic xlink:href="3-20221918_M1.png"/></alternatives></inline-formula>). complementarity sum should constant (marked id="M2">\begin{document}$ u xlink:href="3-20221918_M2.jpg"/><graphic xlink:href="3-20221918_M2.png"/></alternatives></inline-formula>). When dynamic occlusion appears, id="M3">\begin{document}$ xlink:href="3-20221918_M3.jpg"/><graphic xlink:href="3-20221918_M3.png"/></alternatives></inline-formula> fluctuate significantly, so choose id="M4">\begin{document}$ xlink:href="3-20221918_M4.jpg"/><graphic xlink:href="3-20221918_M4.png"/></alternatives></inline-formula> standard judge whether appears. order reduce influence factors (such noise or fluctuation illumination) further removal statistical histogram id="M5">\begin{document}$ xlink:href="3-20221918_M5.jpg"/><graphic xlink:href="3-20221918_M5.png"/></alternatives></inline-formula>. We first find position maximum histogram, then expand both sides histogram. calculate area expanded region, stop expansion greater than threshold. Then id="M6">\begin{document}$ xlink:href="3-20221918_M6.jpg"/><graphic xlink:href="3-20221918_M6.png"/></alternatives></inline-formula> id="M7">\begin{document}$ xlink:href="3-20221918_M7.jpg"/><graphic xlink:href="3-20221918_M7.png"/></alternatives></inline-formula> region measured without interference. Experiments show can retain undisturbed signals significantly improve image. This simple effective, also suitable for general scenes. More importantly, does not need introduce additional verification, which promotes practical process single pixel technology.