An adaptive-optics system usually contains wavefront sensing, wavefront conrection, wavefront reconstructing. So the system is complex and expensive. Deconvolution based on Hartmann-Shark wavefront sensing is a kind of “post-processing”, for it does not need wavefront correction and the system is less complex and less expensive. The principle is sensing wavefront in short exposal and recording blurred images at the same time, which are used in deconvolution. Incremental Wiener filter is first applied to image deconvolution from wavefront sensing. Images of point sources indoors with aberration are processed and restored by incremental Wiener filters. The results show that incremental Wiener filters can efficiently be used in image deconvolution from wavefront sensing. The resolution of restored images can reach a diffraction limit in terms of point sources indoors. Compared with Wiener filter, it extends the selection range of noise restrain coefficient. Even the noise restrain coefficient is not properly chosen, it still has better restored image quality than Wiener filters.