Skip to main content
Fig. 3 | BioMedical Engineering OnLine

Fig. 3

From: A perceptual map for gait symmetry quantification and pathology detection

Fig. 3

a An image where non-subject pixels where naively clipped to zero. The large difference of depth values, existing between the background and the subject, “squeezes” the depth values belonging to the subject and thus causes a decrease in depth resolution in this important and informative part of the image. b An image where background pixels where clipped to the mean (depth) value of subject’s pixels for the whole sequence. Spatial and depth resolution and image details are preserved. c The distribution of the pixel values of the image for the naive clipping (semi-log scale). d The distribution of the pixel values of the image for the smart clipping (semi-log scale). It is important to notice that the whole point of using a default clipping value is to make the distribution of pixel value of the images uni-modal and continuous. This will ensure a well contrasted map for the human eye

Back to article page