Maur Harleen Kaur, Jain Puneet; International Journal of Advance Research, Ideas and Innovations in Technology © 2019, www.IJARIIT.com All Rights Reserved Page | 39 ISSN: 2454-132X Impact factor: 4.295 (Volume 5, Issue 2) Available online at: www.ijariit.com Content based Image Retrieval System using K-Means Clustering Algorithm and SVM Classifier Technique Harleen Kaur Maur harleenkaurmaur@gmail.com Adesh Institute of Engineering and Technology, Faridkot, Punjab Puneet Jain puneetjain@gmail.com Adesh Institute of Engineering and Technology, Faridkot, Punjab ABSTRACT The dramatic rise in the sizes of pictures databases has blended the advancement of powerful and productive recovery frameworks. The improvement of these frameworks began with recovering pictures utilizing printed implications however later presented picture recovery dependent on the substance. This came to be known as Content-Based Image Retrieval or CBIR. Frameworks utilizing CBIR recover pictures dependent on visual highlights, for example, surface, shading and shape, rather than relying upon picture depictions or printed ordering. In the proposed work we will use various types of image features like colour, texture, shape, energy, amplitude and cluster distance to classify the images according to the query image. We will use multi-SVM technique along with the clustering technique to compare the features of the input image with the input dataset of images to extract the similar images as that of the query image. Keywords— CBIR, SVM, Content Based Image Retrieval, Modified SVM, Clustering based SVM Technique 1. INTRODUCTION Retrieval of the relevant images according to the query image from large datasets is becoming more and more challenging today as large collections of images are available today to the public, from image collection to web pages, or even video databases. In recent years, the image retrieval has become an interesting research field due to the use of Image retrieval in various fields like image forensics, criminal investigation system etc [1]. CBIR system has drawn more attention in this field as CBIR aims at developing new techniques for the retrieval of the similar images from a large image dataset by identifying the image contents like colour, texture, shape, intensity, energy etc. In the past, researchers have used various techniques for image retrieval under CBIR like semantic retrieval, relevance feedback, iterative learning and other query methods. The CBIR problem is inspired by the increasing space of image and image databases effectively. For the feature extraction and selection techniques in CBIR the visual content of a still image is used to search the relevant images in large datasets. In general the retrieval process works in two steps, first one is feature extraction step in which the features of every image is extracted and stored in temporary location. The feature describes the contents of the image. Most commonly Visual features used for describing an image are shape , colour, texture, energy , cluster distance etc [5]. The second step which is also termed as classification step compares the features of query image with that of database images and sort images according to the similarity and extracts the result for the user. So the important part of CBIR is development of efficient and effective feature extraction methods. As the example shown below our query image is an image of a dinosaur so the relevant images we search for will be the same as the query images shown in figure. If our query image is: Then the relevant images will be: 1.1 The basic approach for the CBIR system Basically, in the CBIR system, the features of the query image are compared with the features of the images present in the database that were extracted before. After comparison, we are left with the relevant images that match with the query image.