作者: Nils O. Krahnstoever , Jens Rittscher , Gianfranco Doretto , Peter H. Tu
DOI:
关键词:
摘要: A device and method for detecting targets of interest in an image, such as people or objects a certain type. Targets are detected based on optimized strong classifier descriptor that can be combination weak descriptors. The descriptors include user-defined is defined by user to represent shape appearance attribute characteristic parts the target interest. selecting subset exhibit improved performance training images.