Clothing Image Retrieval Based On Multiple Features for Smarter Shopping

Modern life is truly fast paced and lives of most people are overburdened. In such a scenario online shopping is a great and time saver. Ladies clothing cannot be easily specified like grocery or furniture items. Normally, ladies clothing has numerous characteristics that are hard to describe like texture, shape, color, print, length etc. In this work, we propose a way to search for clothes where the query is in the form of image in place of descriptive set of words. The first step of the procedure is to identify in accordance with the length of the dress and sleeves. Next features like color and texture are obtained. To detect the best close match, human intervention is not obligatory. A data set of 1500 images is created. The dataset is built up from craftsvilla, jabong, voonik, myntra, amazon, snapdeal, flipkart, fashionara, shoppersstop. The outcomes confirm a precision of 89.25% and recall of 87.00%.