Shape Based Detection and Top-Down Delineation Using Image Segments

Abstract

We introduce a segmentation-based detection and top-down figure-ground delineation algorithm. Unlike common methods which use appearance for detection, our method relies primarily on the shape of objects as is reflected by their bottom-up segmentation.Our algorithm receives as input an image, along with its bottom-up hierarchical segmentation. The shape of each segment is then described both by its significant boundary sections and by regional, dense orientation information derived from the segment's shape using the Poisson equation. Our method then examines multiple, overlapping segmentation hypotheses, using their shape and color, in an attempt to find a "coherent whole," i.e., a collection of segments that consistently vote for an object at a single location in the image. Once an object is detected, we propose a novel pixellevel top-down figure-ground segmentation by "competitive coverage" process to accurately delineate the boundaries of the object. In this process, given a particular detection hypothesis, we let the voting segments compete for interpreting (covering) each of the semantic parts of an object. Incorporating competition in the process allows us to resolve ambiguities that arise when two different regions are matched to the same object part and to discard nearby false regions that participated in the voting process.We provide quantitative and qualitative experimental results on challenging datasets. These experiments demonstrate that our method can accurately detect and segment objects with complex shapes, obtaining results comparable to those of existing state of the art methods. Moreover, our method allows us to simultaneously detect multiple instances of class objects in images and to cope with challenging types of occlusions such as occlusions by a bar of varying size or by another object of the same class, that are difficult to handle with other existing class-specific top-down segmentation methods.

Symmetry and similarity properties of 2D figures

“…Effective sorting and seeking (e. g. faulty seeking) algorithms are applied in the most different fields as for example in the data sorting [21], [22], in the sound technique [23] and in the image technique [24][25][26][27][28][29][30][31][32][33]. In the image technique one of the most relevant problems is the shape detection consisting of two different partial problems: the shape detection as boundary detection of the image segmentation and the shape detection (shape recognition) as making differences between figures having different shape.…”

“…In the image technique one of the most relevant problems is the shape detection consisting of two different partial problems: the shape detection as boundary detection of the image segmentation and the shape detection (shape recognition) as making differences between figures having different shape. In the case of the boundary detection a lot of effective algorithms are known [24][25][26][27][28]. In the field of the shape recognition the solutions of the most different special problems are worked out as for example a fruit sorting algorithm [29], a face detection algorithm [30], a traffic sign shape classification algorithm [31].…”

Object detection via boundary structure segmentation

“…Segment shape descriptors have been used by [10] for detection and segmentation. Leibe et al [13] combine recognition and segmentation in a probabilistic framework.…”

“…Since T = tt T , this has the meaning that the elements of t are equal to their squared values, which is true only if they are 0 or 1. Finally, the coupling constraints (10) and (11), one of which is quadratic, naturally translate to linear constraints (13) and (14).…”

Automatic object extraction in single-concept videos

“…pedestrians). Gorelick and Basri [5] collected a set of object silhouette exemplars. To extract the object of interest, the authors over-segmented the input image and determined the segments which best matched the associated templates.…”