An optimized PCNN for image classification

In recent years, Image classification has been a growing research area in the computer vision field. Thus, many approaches were proposed in literature. Moreover, many content-based image classification approaches are widely used in developing applications and techniques for many areas such as remote-sensing and content-based image retrieval. In this study, we introduce a new technique for content-based image classification. This technique combines the Pulse-Coupled Neural Network (PCNN) with K-Nearest Neighbors (K-nn) for image classification. The PCNN is used as feature extractor that extracts the visual feature of the images as signatures. Afterwards, the K-nn work as a classifier that process these signatures by examining the distance between them to detect the image class. Furthermore, we implemented prototype to validate our technique. The presented results show that the proposed approach can classify images efficiently.