Object Recognition Based on the Context Aware Decision-Level Fusion in Multiviews Imagery

摘要

Spectral similarities and spatial adjacencies between various kinds of objects, shadow, and occluded areas behind high-rise objects as well as the complex relationships between various object types lead to the difficulties and ambiguities in object recognition in urban areas. Using a knowledge base containing the contextual information together with the multiviews imagery may improve the object recognition results in such a situation. The proposed object recognition strategy in this paper has two main stages: single view and multiviews processes. In the single view process, defining region’s properties for each of the segmented regions, the object-based image analysis (OBIA) is performed independently on the individual views. In the second stage, the classified objects of all views are fused together through a decision-level fusion based on the scene contextual information in order to refine the classification results. Sensory information, analyzing visibility maps, height, and the structural characteristics of the multiviews classified objects define the scene contextual information. Evaluation of the capabilities of the proposed context aware object recognition methodology is performed on two datasets: 1) multiangular Worldview-2 satellite images over Rio de Janeiro in Brazil and 2) multiviews digital modular camera (DMC) aerial images over a complex urban area in Germany. The obtained results represent that using the contextual information together with a decision-level fusion of multiviews, the object recognition difficulties and ambiguities are decreased and the overall accuracy and the kappa are gradually improved for both of the WorldView-2 and the DMC datasets.