Image Retrieval with Fusion of Thepade’s Sorted Block Truncation Coding n-ary based Color and Local Binary Pattern based Texture Features with Different Color Places
Sudeep D. Thepade1, Rohan Awhad2, Prakhar Khandelwal3

1Sudeep D. Thepade*, Computer Engineering Department, Pimpri Chinchwad College of Engineering, Pune, India.
2Rohan Awhad, Computer Engineering Department, Pimpri Chinchwad College of Engineering, Pune, India.
3Prakhar Khandelwal, Computer Engineering Department, Pimpri Chinchwad College of Engineering, Pune, India.
Manuscript received on February 10, 2020. | Revised Manuscript received on February 20, 2020. | Manuscript published on March 10, 2020. | PP: 28-34 | Volume-9 Issue-5, March 2020. | Retrieval Number: E1963039520/2020©BEIESP | DOI: 10.35940/ijitee.E1963.039520
Open Access | Ethics and Policies | Cite | Mendeley
© The Authors. Blue Eyes Intelligence Engineering and Sciences Publication (BEIESP). This is an open access article under the CC BY-NC-ND license (

Abstract: In these years, there has been a gigantic growth in the generation of data. Innovations such as the Internet, social media and smart phones are the facilitators of this information boom. Since ancient times images were treated as an effective mode of communication. Even today most of the data generated is image data. The technology for capturing, storing and transferring images is well developed but efficient image retrieval is still a primitive area of research. Content Based Image Retrieval (CBIR) is one such area where lot of research is still going on. CBIR systems rely on three aspects of the image content namely texture, shape and color. Application specific CBIR systems are effective whereas Generic CBIR systems are being explored. Previously, descriptors are used to extract shape, color or texture content features, but the effect of using more than one descriptor is under research and may yield better results. The paper presents the fusion of TSBTC n-ary (Thepade’s Sorted n-ary Block Truncation Coding) Global Color Features and Local Binary Pattern (LBP) Local Texture Features in Content Based Image with Different Color Places TSBTC n-ary devises global color features from an image. It is a faster and better technique compared to Block Truncation Coding. It is also rotation and scale invariant. When applied on an image TSBTC n-ary gives a feature vector based on the color space, if TSBTC n-ary is applied on the obtained LBP (Local Binary Patterns) of the image color planes, the feature vector obtained is be based on local texture content. Along with RGB, the Luminance chromaticity color space like YCbCr and Kekre’s LUV are also used in experimentation of proposed CBIR techniques. Wang dataset has been used for exploration of proposed method. It consists of 1000 images (10 categories having 100 images each). Obtained results have shown performance improvement using fusion of BTC extracted global color features and local texture features extracted with TSBTC n-ary applied on Local Binary Patterns (LBP). 
Keywords: CBIR, Kekre’s LUV, TSBTC n-ary, Wang Dataset
Scope of the Article: Network coding