SYSTEM OR METHOD FOR CLASSIFYING IMAGES

摘要

A system or method is disclosed for classifying sensor images into one of several pre-defined classifications (22). Mathematical moments (72) relating to various features (74) or attributes in the sensor image (26) are used to populated a vector of attributes (28), which are then compared to a corresponding template vector of attribute values (87). The template vector (87) contains values for known classifications (22) which are preferably predefined. By comparing the two vectors, various votes (92) and confidence metrics (85) are used to ultimately selct the appropriate classification (22). In some embodiments, preparation processing is performed before loading the attribute vector (87) with values. Image segmentation (69) is often desirable. The performance of heuristics (73) to adjust for environmental factors such as lighting can also be desirable. One embodiment of the system (29) is to prevent the deployment of an airbag when the occupant (34) in the seat (36) is a child, a rearfacing infant seat (36), or when the seat (36) is empty.