Multisensor Integration

This paper presents a system f o r classification of scenes using a multisensor integration framework. Indoor scenes are imaged using a visual and a n infrared sensor and the images processed in three stages to perf o r m classification of sensed objects into two classes: human and background. Finally, information f rom individual classifiers is integrated in order to obtain an improved classification performance. Details of feature extraction and classification using neural network combining are presented. Segmentation of the imaged scene is performed using existing techniques such as texture analysis and histogram modeling. Classification results on real-world data are presented. The syst e m represents a first step in the development of improved, robust classifiers based on the concepts of neural networks and multisensor integration.