Content-Based Image Retrieval for Super-Resolution Images using Feature Fusion: DL& Hand Crafted

TO PURCHASE OUR PROJECTS IN ONLINE (OR) OFFLINE
CONTACT:VENKAT INNOVATIVE PROJECTS
NAME: VENKATARAO GANIPISETTY
Mobile & WhatsApp :+91 9966499110
Mobile & WhatsApp :+91 9573201550
Email :[email protected]
Email :[email protected]
website:https://venkatinnovativeprojects.com/
About Project:
In propose paper author combining multiple techniques such as deep learning High level and Hand crafted features extraction from Super Resolution images to retrieve content based similar images. All existing algorithms were utilizing either deep learning or manual hand crafted features extraction technique whose retrieval precision rate is very low. In propose paper author validated propose model with multiple different techniques on different dataset and in all propose model outperform.
In propose paper InceptionV3 GOOGLENET model utilize to extract high level features and then employ hand crafted features extraction techniques called Modified DDBTC, HOG, Integrated channel HSI and GCLM. Both High level and hand crafted features will get concatenated and then employ Vector Similarity search algorithm to retrieve content based similar images.
Author validated propose model with multiple techniques and it’s not possible to implement all so we have utilize GOOGLENET features as existing and combination of GOOGLENET + Hand crafted as propose model. We have trained both existing and propose model on VISTEX and STEX dataset.
Both dataset images are converted to super solution images by employing INTER-CUBIC algorithm.