Isotropic and anisotropic edge detection based on Fourier single pixel imaging

In this paper, we propose a novel edge detection scheme based on Fourier single pixel imaging, capable of realizing both isotropic and anisotropic edge detection. In the scheme, we use a series of specific designed sinusoidal speckle patterns to illuminate the object. The corresponding total intensity of reflective light is then measured by a bucket detector. By performing the inverse Fourier transformation on the bucket detection results, an isotropic or anisotropic edge image of the object could be reconstructed directly. Simulation and experimental results demonstrate that our scheme can capture the isotropic or anisotropic edge images of the object with both direct and indirect line of sight to the object. Our scheme offers the advantages of achieving a clear and high-quality isotropic or anisotropic edge image with only one imaging reconstruction computing operation from sub-Nyquist measurements, which significantly reduces the number of measurements required for single pixel imaging. In addition, our scheme provides the benefits of achieving the single pixel anisotropic edge imaging, allowing one to decide the edges in an arbitrary direction to disappear while preserving the edges in other directions.