Enhancing Hyperspectral Image Quality for Chicken Eggs Using Unmixing Approach

The current non-destructive and fast method of hyperspectral imaging technology for the chicken egg-related problems in agri-food processing suffers from the quality of acquired hyperspectral images. Therefore, the results of all existing attempts to deal with those egg-related problems (e.g., freshness, grading eggs, fertility detection and distinguishing abnormal eggs) can be improved since they are highly related to the quality of hyperspectral images taken over eggs. This project aims to address these challenges and present a hyperspectral imaging enhancement scheme to accurately determine the spectral profile of internal components of egg (e.g., yolk, egg white, eggshell) using one of the strong remote sensing tools called spectral unmixing. Eventually, the quality of hyperspectral image of eggs will be enhanced based on unmixing approach along with better estimating the noise level and appropriate denoising technique.

Faculty Supervisor:

Saeed Gazor

Student:

Yaser Esmaeili Salehani

Partner:

MatrixSpec Solutions

Discipline:

Engineering - computer / electrical

Sector:

Agriculture

University:

Queen's University