Holycious/Fruit-Classifier-Kel-3

🍎 Fruit Classification Model

Classify fruits into 5 categories: Apple, Apricot, Nectarine, Peach, and Pear.

Model Architecture

• —Algorithm: Support Vector Machine (SVM) with RBF kernel
• —Features: 79-dimensional feature vector
• —64 RGB histogram features
• —7 Unser texture features
• —8 shape features
• —Preprocessing: Automatic fruit segmentation and cropping
• —Performance:
• —Test Accuracy: ~95%
• —Cross-validation: 5-fold CV

How to Use

1. 1.Upload a clear image of a fruit
2. 2.The model automatically preprocesses and extracts features
3. 3.View prediction results with confidence scores

Technical Details

The model uses traditional computer vision and machine learning techniques:

• —Image segmentation using adaptive thresholding
• —Handcrafted feature extraction
• —SVM classification with hyperparameter tuning
• —PCA for dimensionality reduction

Dataset

Trained on a diverse dataset of pome and drupe fruits with multiple varieties.

Credits

Developed using Scikit-learn, OpenCV, and Gradio.