Skip to main content
Fig. 6 | IPSJ Transactions on Computer Vision and Applications

Fig. 6

From: Sample-based integrated background subtraction and shadow detection

Fig. 6

CDnet shadow detection results shadow detection results for four frames from the CDnet shadow category. On the first row is the input frame, on the second row is the ground truth, and on the third row is the result of the presented background subtraction method. Pixels outside the region of interest are assigned a gray value of 85 in ground truth. On the fourth row, pixels in the initial mask M1 are assigned a gray value of 127. On the last row, pixels in the final shadow mask M2 are assigned a gray value of 50, as in the ground truth. It can be seen that many pixels which pass the initial property test are not labeled as shadow in the final mask M2 due to insufficient matches with the samples in the shadow model. This illustrates the usefulness of shadow modeling with samples. a Cubicle. b Bungalows. c People in shade. d Bus station

Back to article page