Graph Cut Based approach in creating superpixel for Depth estimation
VaibhavWaman, Dr.Rahul M Mulajkar · International journal of advance research and innovative ideas in education · 2018
The Target of estimating the depth of each pixel in a scene from a single image which is a monocular image using the monocular cues. In past research work approaches [1, 2], which attempt to map from appearance features to depth directly, we first perform a generate segmentation (creation of Superpixels) of the scene and use the semantic labels to guide the 3D reconstruction. This approach provides several advantages: By knowing the segmentation class of a pixel or region, depth and geometry constraints can be easily enforced ( which gives idea about “sky” is far away and “ground” location of scene). In addition, depth can be more readily predicted by measuring the difference in appearance with respect to a given segmentation class. For example, a tree will have more uniform appearance in the distance than it does close up.by means of which, the incorporation of segmentation features allows us to achieve state-of-the-art results with a significantly simple model than previous research works.various models have been used to convert the monocular image into the 3d reconstruction model.