BLACKBEANS CURRY

Dataset Description: Black Bean (Cicer arietinum) Good and Bad Classification

1. Overview

This dataset consists of images of black beans (Cicer arietinum) captured for classification into "Good" and "Bad" categories. The dataset is designed for automated quality assessment based on visual properties. The images were collected under standardized conditions to ensure consistency and accuracy in classification.

2. Data Collection

Total Samples: 500+ black beans (Good and Bad).

Camera Used: iPhone 15 mobile camera.

Lighting Conditions: Natural daylight.

Background: White background for clear contrast.

3. Image Characteristics

Resolution: High-resolution images ensuring detailed visibility of bean surface texture.

Orientation: Beans placed flat, allowing clear identification of surface features.

Focus: Sharp images highlighting key attributes such as color, texture, and defects.

Color Representation: The dataset captures variations in black shades, distinguishing fresh (good) beans from aged or damaged (bad) ones.

4. Classification Criteria

Good Black Beans:

Deep black or dark brown color with a smooth, glossy surface.

Uniform size and shape without cracks.

Hard texture, indicating freshness.

Bad Black Beans:

Discoloration, including faded, dull, or reddish shades.

Wrinkles, cracks, or damaged outer shells.

Mold, insect damage, or shriveled appearance.

Irregular shape or size variations.

5. Application and Use Cases

Machine learning and deep learning models for automated quality classification.

Quality control in agriculture, food processing, and storage.

AI-based grading systems for commercial black bean sorting.

Research on bean spoilage and preservation techniques.

6. Potential Preprocessing Steps

Image Resizing: Standardizing dimensions for model training.

Augmentation: Adjusting brightness, contrast, and rotation for dataset robustness.

Background Processing: Ensuring focus on beans by segmenting out the white background.

Feature Extraction: Analyzing color, shape, and texture for enhanced classification accuracy.

This dataset provides a valuable foundation for AI-driven agricultural quality assessment, enabling automated classification of good and bad black beans with precision.