Ghost imaging in convolution space using shifted random-pattern illumination

We present a ghost imaging technique that enables accurate acquisition of convolution images. Conventional speckle-shifting ghost imaging achieves convolution-space imaging indirectly but requires multiple pattern shifts for each image update, which reduces measurement efficiency. The proposed approach addresses this limitation by employing shifted random-pattern illumination, which approximates ghost imaging in the convolution space while improving efficiency. Both theoretical analysis and experimental results demonstrate that the proposed method achieves higher image quality than speckle-shifting ghost imaging. Furthermore, experimental validation in X-ray ghost imaging highlights the potential of the proposed approach for future high-speed imaging applications in invisible-light regimes.