sahanave/womenwhocodehsidemo

Hyperspectral Image(HSI) Analysis is one of the cutting-edge fields in Artificial Intelligence(AI) research due to its applications in various fields from agriculture to surveillance. This repo is to assist beginners to get started with processing HSI images focusing on data ingestion and data visualization.

WomenWhoCodeHSIDemo

Objective

Hyperspectral Image(HSI) Analysis is one of the cutting-edge fields in Artificial Intelligence(AI) research due to its applications in various fields from agriculture to surveillance. This repo is to assist beginners to get started with processing HSI images focusing on data ingestion and data visualization. This repo was designed as a demo for my talk on Hyperspectral image analysis during WomenWhoCode Open-house talk.

Task

Object detection using supervised or unsupervised approaches on colour images is a huge field. Recently, a few studies demonstrated that efficient salient object detection can also be implemented by using visible spectrum of hyperspectral images. We want to investigate what kind of information exists in hyperspectral space for object detection.

Background

A hyperspectral salient object detection dataset with a collection of 60 hyperspectral images with their respective ground-truth binary images and corresponding colour images (sRGB).

Wavelength range : 350 - 1100 nm (81 bands)

Objects : (no description in the paper) : Foreground

Dataset

Dataset : https://arxiv.org/abs/1806.11314