Akasha, N, M et al. Int. Journal of Engineering Research and Application www.ijera.com ISSN : 2248-9622, Vol. 7, Issue 2, ( Part -2) February 2017, pp.55-60 www.ijera.com DOI: 10.9790/9622- 0702025560 55 | Page Logo Matching for Document Image Retrieval Using SIFT Descriptors Boyapati Bharathidevi 1 , Lakshmi Prasad Chennamsetty 2 , Ammeti Raghava Prasad 3 , Ashok Kumar Balijepalli 4 1 Asst.Professor, Dept of ECE, Universal College Of Engineering & Technology, Perecherla, Guntur(dt), A.P, India-522438 2 Asst.Professor, Dept of ECE, Universal College Of Engineering & Technology, Perecherla, Guntur(dt), A.P, India-522438. 3 Asst.Professor,Dept of ECE, Universal College Of Engineering & Technology, Perecherla, Guntur(dt), A.P, India-522438 4 Asst.Professor,Dept of ECE, Universal College Of Engineering & Technology, Perecherla, Guntur(dt), A.P, India-522438 ABSTRACT In current trends the logos are playing a vital role in industrial and all commercial applications. Fundamentally the logo is defined as it’s a graphic entity which contains colors textures, shapes and text etc., which is organized in some special visible format. But unfortunately it is very difficult thing to save their brand logos from duplicates. In practical world there are several systems available for logo reorganization and detection with different kinds of requirements. Two dimensional global descriptors are used for logo matching and reorganization. The concept of Shape descriptors based on Shape context and the global descriptors are based on the logo contours. There is an algorithm which is implemented for logo detection is based on partial spatial context and spatial spectral saliency (SSS). The SSS is able to keep away from the confusion effect of background and also speed up the process of logo detection. All such methods are useful only when the logo is visible completely without noise and not subjected to change. We contribute, through this paper, to the design of a novel variation framework able to match and recognize multiple instances of multiple reference logos in image archives. Reference logos and test images are seen as constellations of local features (interest points, regions, etc.) and matched by minimizing an energy function mixing: 1) a fidelity term that measures the quality of feature matching, 2) a neighborhood criterion that captures feature co-occurrence geometry, and 3) a regularization term that controls the smoothness of the matching solution. We also introduce a detection/recognition procedure and study its theoretical consistency. Index Terms: Context-dependent kernel, logo detection, logo recognition. I. I.INTRODUCTION Logos are often used pervasively as declaration of document source and ownership in business and government documents. The problem of logo detection and recognition is of great interest to the document image analysis and retrieval communities because it enables immediate identification of the source of documents based on the originating organization. Logos are graphic productions that either recall some real world objects, or emphasize a name, or simply display some abstract signs that have strong perceptual appeal [see Fig. 1(a)]. Color may have some relevance to assess the logo identity. But the distinctiveness of logos is more often given by a few details carefully studied by graphic designers, semiologists and experts of social communication. The graphic layout is equally important to attract the attention of the customer and convey the message appropriately and permanently. Fig. 1. (a) Examples of popular logos depicting real world objects, text, graphic signs, and complex layouts with graphic details. (b) Pairs of logos with malicious small changes in details or spatial arrangements. (c) Examples of logos displayed in real world images in bad light conditions, with partial occlusions and deformations. RESEARCH ARTICLE OPEN ACCESS