Knowledge acquisition for language description from scene understanding

The motivation behind this work is to explore and pursue benefits of Cognitive Informative and Artificial Intelligence. Computer based Image Analysis and Machine Vision are based on Image Processing Algorithms. Choosing the best concept to comprehend a meaning in natural language generation is one of the tough tasks. A vision system while navigating in an environment should be able to recognize what the main objects in the scene are. Here, we are proposing a system, which automatically generates natural language descriptions from scene i.e. images. It acquires knowledge from the scene using theories of cognitive science. Spatial and semantic relations are used to produce a knowledge base and subsequently it transcribes the knowledge in a natural language form.