Abstract
Moving objects often appear normal or even sharper than they actually are (motion sharpening). Earlier we found dramatic motion sharpening even when more than half of the frames of a moving image were greatly blurred (Takeuchi & De Valois, VSS, 2002). Now we have further examined the sharpening process by varying the spatial frequency components of moving natural images. We prepared movies from natural images. Each contained 100 frames. All frames were blurred by low-pass filtering then combined to make a movie of 30 fps. Subjects judged the perceived sharpness of the movies by comparing each to a still image. On each trial, subjects adjusted three parameters of the still image; overall luminance contrast, the slope of the amplitude in the spatial frequency domain, and cut-off spatial frequency. We found that subjects mainly varied the slope of the amplitude to match the perceived sharpness between the moving image and the still image. In the next experiment, we used moving images in which each frame was modulated by low-pass, high-pass, or band-reject filters. We found the strongest motion sharpening when the image frames were band-reject filtered. Our results indicate that the perceived contrast of higher spatial frequency components in the moving images is enhanced in the motion sharpening process, but lower spatial frequency components are also needed to induce the strongest motion sharpening. This indicates an interaction between different spatial frequency components in the motion sharpening process. We have applied our findings to make blurred (compressed) moving images appear sharper than they actually are. When text was scrolled at high velocity, the percentage of letters correctly identified could be increased if the texts were blurred by changing the slope of the amplitude of the spatial frequency spectrum.