A New Method for Indoor-outdoor Image Classification Using Color Correlated Temperature

In this paper a new method for indoor-outdoor image classification is presented; where theconcept of Color Correlated Temperature is used to extract distinguishing features betweenthe two classes. In this process, using Hue color component, each image is segmented intodifferent color channels and color correlated temperature is calculated for each channel.These values are then incorporated to build the image feature vector. Besides colortemperature values, the feature vector also holds information about the color formation of theimage. In the classification phase, KNN classifier is used to classify images as indoor oroutdoor. Two different datasets are used for test purposes; a collection of images gatheredfrom the internet and a second dataset built by frame extraction from different videosequences from one video capturing device. High classification rate, compared to other stateof the art methods shows the ability of the proposed method for indoor-outdoor imageclassification.