The invention discloses a single-pixel calculation imaging method. The method comprises the following steps: firstly generating a group of random binary square matrix phi 1, phi 2, phi 3, ..., phi R, and sequentially performing inner product operation with normalized images G (i, j) to generate single-pixel data streams Y1, Y2, Y3, ..., YR; performing Fourier transform on the Y1, Y2, Y3, ..., YR, and performing data compression and data restoration on the frequency domain data streams by using the frequency domain data sparsity and the center symmetry; and performing constraint L1 norm optimization calculation by using a random binary matrix phi (L*m) and the restored Y1, Y2, Y3, ..., YR subjected to inverse Fourier transform to reconstruct an original image and realize the purpose of single-pixel calculation imaging. By adoption of the method, a single-pixel camera can be provided to realize a digital simulation process physically, and the difficulty of big data storage, transmission and high energy consumption of aerospace sensors is solved.