Assessment of banana fruit maturity by image processing technique

Area of Science:

• Agricultural Science
• Computer Vision
• Food Science

Background:

• Banana fruit quality and marketability are significantly influenced by maturity at harvest.
• Accurate identification of banana maturity is crucial for optimizing harvesting and preventing under- or over-ripening.

Purpose of the Study:

• To develop and assess an image processing technique for precise detection of fresh banana fruit maturity stages.
• To utilize color and size features from fruit images for maturity classification.

Main Methods:

• Collected 120 images of bananas across three maturity stages (under-mature, mature, over-mature).
• Extracted image features including mean color intensity and size parameters (area, perimeter, major/minor axis length).
• Developed and evaluated two classification algorithms: mean color intensity and area-based algorithms.

Main Results:

• Mean color intensity and area were identified as significant predictors of banana maturity.
• The mean color intensity algorithm achieved 99.1% accuracy in classifying banana fruit maturity.
• The area algorithm demonstrated 85% accuracy in classifying under-mature fruit.

Conclusions:

• Image processing offers a precise and effective method for assessing banana fruit maturity.
• The developed algorithms can support the commercial development of automated field systems for optimal harvest decision-making.
• This technology empowers banana growers to improve harvest timing, leading to enhanced fruit quality and market value.