Scene Categorization by Hierarchical Clustering on Adaptive Spatio-Temporal Features

The complex task of vision based analytics by computer gets more challenging when mutual motion between camera assembly and object of interest creates false motion features in the scene. This adds further complexity when adverse environmental conditions are handled along with scene understanding, scene captioning, object and activity recognition. The current work has proposed a novel hierarchical categorization based methodology to categorize scenes based on selected spatio-temporal features. The procedure of determination of feature hierarchy and the detailed description of step by step methodology have been provided along with comprehensive results.