Litcius/Paper detail

Implementation of Super Resolution in Images Based on Generative Adversarial Network

K. Srinivasa Reddy, Vinodh P Vijayan, Ayan Das Gupta, Prabhdeep Singh, R.G. Vidhya, Dhiraj Kapila

20222022 8th International Conference on Smart Structures and Systems (ICSSS)26 citationsDOI

Abstract

A 3D visualization of a microscopic object is provided by the integral imaging microscopy system. A generative-adversarial-network (GAN) relied on super resolution (SR) algorithm is suggested in this research to improve resolution. The generator in GAN network regresses the highresolution (HR) outcome out of the low-resolution (LR) input image, where the discriminator differentiates among the original as well as generated images. It could perhaps recover the edges and boost the resolution besides 2, 4, or indeed 8 times without compromising image quality for different sector in different field. The framework is validated using a variation of decreased microscopic specimen images as well as appropriately develops images with considerable directional view and compared with each other to get the best model among them in different sector. The quantifiable investigation reveals that the suggested framework outperforms the existing algorithms for microscopic images.

Topics & Concepts

DiscriminatorGenerator (circuit theory)Computer scienceVisualizationArtificial intelligenceAdversarial systemObject (grammar)Resolution (logic)Image (mathematics)Computer visionGenerative grammarGenerative adversarial networkImage resolutionPattern recognition (psychology)DetectorPhysicsTelecommunicationsQuantum mechanicsPower (physics)Advanced Image Processing TechniquesImage Processing Techniques and ApplicationsDigital Holography and Microscopy
Implementation of Super Resolution in Images Based on Generative Adversarial Network | Litcius