Nimisha22/Vegetable-Detection

Vegetable Detection

Problem Statement

Although there are pre-trained models available for detecting vegetables on the ImageNet dataset, it remains unclear whether the accuracy of these models holds true for other datasets. Additionally, these models lack the ability to predict important attributes such as the age, ripeness, damage, unripeness, or dryness of the vegetable.

Dataset

Vegetables:

Tomato
Bell Peper
New Mexico Green Chile
Chile Peper

Vegetables are classified as:

Damaged
Dried
Old
Ripe
Unripe


Accuracy of pre-existing models is calculated on VegNet dataset. Then training and testing is done to detect the vegetable along with its quality.