3D single pixel imaging based on parallel measurement with quadrant detector

Abstract

Structured light three-dimensional (3D) imaging has advantages such as high accuracy, high resolution, and non-contact, and has enormous application value in fields such as automotive manufacturing and cultural relic detection. However, it often requires multiple structured light encoding to obtain 3D information, thus limiting the speed of 3D imaging. Single pixel imaging (SPI) technology, due to its use of structured light and single point detection to jointly obtain image information, can simply achieve simultaneous detection of multi-dimensional information through a single pixel detector array. Therefore, the structured light 3D imaging technology is combined with the single-pixel technology of multi-channel quadrant sensing, and the modulation of three structured light fields of red, green, and blue light is achieved separately through the decoupling of spatial 3D information and spectral dimension information. Combined with a quadrant sensing detector integrated with red, green, and blue filtering, simultaneous measurement of three structured light field signals is achieved. Thus, a scheme demonstration is accomplished to improve the imaging speed of 3D imaging by three times through decoupling. Further combining Gray codes and optimizing Hadamard sequences using compressive sensing ensures the accuracy and imaging quality under undersampling of 3D imaging. The experimental results show that the RMSE of our method is only 0.0576 mm. This method can be further extended to achieve high-precision and high-quality 3D reconstruction using more channel structured light modulation and more spectral detector arrays in only one parallel measurement.