Detecting, Counting Objects to Form Color Image Features

Abstract: Color image features are very important victor of values. Which is used as an image identifier or primary key which can be used easily in a color image retrieval system, the features victor can be used also as an image signature to compute image classifier in a color image recognition system.  In this paper research we will introduce a texture method to create color image features. This method will have based on using crossing number to detect some objects in the image, the detected objects where isolated, ending, connected and crossing points. The idea of the proposed method is to find the count of each objects, these counts will form the image features. The proposed method will be implemented and the experimental results will be analyzed to prove the efficiency, accuracy, and stability of the proposed method.

Keywords: Features, crossing number, isolated point, ending point, connected point, crossing point.