Fruit Battery Method for Oil Palm Fruit Ripeness Sensor and Comparison with Computer Vision Method.

Area of Science:

• Agricultural Science
• Electrical Engineering
• Computer Vision

Background:

• Traditional oil palm fresh fruit bunch (FFB) ripeness evaluation relies on human vision, which can be inaccurate.
• Inaccurate ripeness grading leads to reduced oil palm fruit oil extraction rate (OER).
• Objective and accurate methods are needed to determine FFB ripeness for optimal OER.

Purpose of the Study:

• To evaluate the fruit battery method for distinguishing oil palm FFB ripeness stages using load resistance voltage and moisture content.
• To compare the accuracy of the fruit battery method with computer vision (using support vector machine - SVM) and a combined approach.
• To identify the most accurate method for FFB ripeness assessment to enhance OER.

Main Methods:

• Developed a fruit battery method measuring load resistance voltage and moisture content for FFB ripeness.
• Utilized computer vision analyzing color features with a support vector machine (SVM) for ripeness classification.
• Tested and compared the accuracy scores of the fruit battery method, computer vision, and a hybrid approach.

Main Results:

• A 1 kΩ load resistance demonstrated the best moisture content resolution in the fruit battery method.
• The combined fruit battery and computer vision method achieved the highest accuracy score for FFB ripeness evaluation.
• The fruit battery method showed higher accuracy than computer vision alone.

Conclusions:

• The combined fruit battery and computer vision approach offers superior accuracy for oil palm FFB ripeness assessment.
• The fruit battery method, particularly at 1 kΩ load resistance, provides a reliable measure of FFB ripeness and moisture content.
• Implementing these advanced methods can significantly improve oil extraction rates by ensuring optimal FFB ripeness grading.