Scene classification of ambiguous visual information

A framework for scene classification of ambiguous visual information is described and validated. Context-based scene classification algorithm is used for ambiguous visual information applications. The system can differentiate ambiguous visual information from various semantic meanings using a spatial layout of context information, which capture the “essential” of the scene. Distinct from previous frameworks, the system presents the entire scheme of being biologically plausible and application efficiency, offering a straightforward platform for rapid analysis and interpretation on ambiguous visual information demonstrating generalization and scalability of the approach.