GrCS: Granular Computing-Based Crowd Segmentation

Abstract

Crowd segmentation is important in serving as the basis for a wide range of crowd analysis tasks such as density estimation and behavior understanding. However, due to interocclusions, perspective distortion, clutter background, and random crowd distribution, localizing crowd segments is technically a very challenging task. This paper proposes a novel crowd segmentation framework-based on granular computing (GrCS) to enable the problem of crowd segmentation to be conceptualized at different levels of granularity, and to map problems into computationally tractable subproblems. It shows that by exploiting the correlation among pixel granules, we are able to aggregate structurally similar pixels into meaningful atomic structure granules. This is useful in outlining natural boundaries between crowd and background (i.e., noncrowd) regions. From the structure granules, we infer the crowd and background regions by granular information classification. GrCS is scene-independent and can be applied effectively to crowd scenes with a variety of physical layout and crowdedness. Extensive experiments have been conducted on hundreds of real and synthetic crowd scenes. The results demonstrate that by exploiting the correlation among granules, we can outline the natural boundaries of structurally similar crowd and background regions necessary for crowd segmentation.

A hybrid social influence model for pedestrian motion segmentation

“…the granular crowd segmentation (GCS) [17], and collective transition (CT) [37]. Our experimental results show that the HSIM achieves superior pedestrian motion segmentation.…”

“…For this purpose, the nite time lyapunov exponent [36] is exploited to dene the boundaries of dierent segments. Kok et al [17] propose the granular computing to aggregate similar pixels into atomic structure granules for motion segmentation. The structure granules are used to isolate the motion and background regions.…”

“…2. Our proposed method is driven by low computational overheads compared to the state-of-the-art methods [2][14] [39][27] [17][37]. 3.…”

“…Our experimental results show that the HSIM achieves superior pedestrian motion segmentation. For quantitative analysis, we compute both the Jaccard Similarity Metric (JSM) [38] and the F-score [17]. Our proposed HSIM outperforms all the reference methods in both qualitative and quantitative evaluations.…”

Attribute-Driven Granular Model for EMG-Based Pinch and Fingertip Force Grand Recognition

“…Granular computing (GrC) solves problems via making use of granules, i.e. groups, classes or clusters of a universe, which is closely related to the cognitive strategy of human being in problem solving and it is technically transferable to the design of human-centric intelligent systems [17], which has been applied in image-based crowd segmentation [18], longterm prediction model for the energy system [19], [20], video based object tracking [21] and principle curve extraction [22], etc.…”