RESEARCH PROJECT 03
FOOD QUALITY
Artificial intelligence and machine learning research for food quality detection, including fruit image classification and milk adulteration analysis.
FOOD QUALITY
Computer vision meets food quality analysis.RESEARCH OVERVIEW
AI-Based
Food Quality
Detection
The research investigated machine learning and deep learning approaches for food-quality detection.
The documented work included fruit image collection and classification, together with an experiment involving classification of raw and adulterated milk samples.
EXPERIMENT 01
Fruit Image
Classification
Dataset Collection
Images of fruits such as apples and oranges were collected from different fruit vendors.
The images included different viewing angles and backgrounds.
CNN Model
A convolutional neural network was developed for fruit classification using the collected image dataset.
A graphical user interface was also documented for the fruit classifier.
EXPERIMENT 02
Milk
Adulteration
The documented milk experiment evaluated machine learning classifiers using a dataset containing raw and adulterated milk samples.
CLASSIFICATION RESULTS
Machine Learning
Models
FOOD QUALITY FEATURES
What Can Be
Measured?
Colour
Visual colour information used for quality and maturity analysis.
Shape & Size
Morphological characteristics associated with fruit quality.
Texture
Surface and texture characteristics for image-based analysis.
Composition
Parameters such as temperature, moisture, pH and other measurements.
PROPOSED RESEARCH DIRECTION
Visual +
Volatile
Features
The research material proposed combining visual characteristics with volatile compounds and other measurable characteristics.
Proposed inputs included visual features such as colour, shape, size and texture, together with compounds such as ethylene, carbon dioxide and oxygen.
Machine learning and deep learning models could then be trained to predict fruit ripeness and quality.