关键词: 图像识别, 重要度, 知识域, 特征向量, 权值

Abstract: Importance is a newer method of computation in rough set theory and reflects the degree to which an attribute affects a domain.An image recognition method based on importance theory was proposed to improve the accuracy of image recognition and ensure the timeliness of algorithm calculation.The definition and theorem of importance theory was given,and then the image feature knowledge domain was built according to the mapping of image features.On this basis,the importance of image feature knowledge was calculated.The importance of image feature knowledge was taken as the weight of image feature vector calculation to identify the image.Simulation results showed that the proposed method could effectively identify the image,and the algorithm had high computational efficiency.

Key words: image recognition, importance, knowledge domain, feature vector, weight