Abhinav Narayan

Compared image registration models - SURF, Scale Invariant Feature Transform(SIFT), and Binary Robust Invariant Scalable Keypoints (BRISK). Designed and developed an invisible pattern extraction system for both images and video. Conducted a research study on image fusion using wavelet transform to enhance low light images. Implemented image quality metrics to the image registration models to find genuine products and fake products. Converted MATLAB algorithms to Python - Adaptive Gamma Correction, Adaptive Switching Mean Filter. Developed a machine learning system using clustering to classify good and bad quality images.

Designed and developed image registration system using Speeded Up Robust Features(SURF). Conducted a research study on image denoising using undecimated wavelets.

Worked with Professor Prakash Tunga P and a few versatile individuals on the publication ”Automatic Brain Tumor Segmentation using Dense-net.” MRI scans were taken from BRATS 2017/18 dataset. These scans were trained and tested using Densely connected convolutional neural network(Dense-net). These images were compared with the ground truth to avoid misdiagnosis of patient. This model was designed and developed using Python software, using Keras and TensorFlow libraries.

Using basics of LabVIEW, a file installer and uninstaller was made for NI-myRIO.

Karnataka, India-560098