Scholarlink

Accurate classification of Palm Oil Fresh Fruit Bunch (POFFB) ripeness is crucial for optimizing yield and quality in the palm oil industry. Traditional manual methods are time-consuming and prone to error. This study presents an integrated approach utilizing Near-Infrared (NIR) optical sensors combined with artificial data generation and Case-Based Reasoning (CBR) to enhance classification accuracy and efficiency. NIR optical sensors capture spectral color data indicative of POFFB ripeness levels while artificial data generation is employed to provide comprehensive datasets overcoming the limitations and cost implications of preparing numerous real POFFB samples. The CBR method is used for classification and evaluated using performance metrics. The result shows that the classification model demonstrates a strong performance with an overall accuracy of 98.93%, sensitivity and precision both at 97.32%, and specificity at 99.33%. The study reveals the strong potential of these approaches to enhance efficiency and reliability in POFFB ripeness classification.

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Integrated Approaches for Palm Oil Fresh Fruit Bunch Ripeness Classification Using Near-Infrared Optical Sensors, Artificial Data and Case-Based Reasoning

  • 会议论文

  • 发表日期：

    2026年

查看全文
  • Aidil Shafiza Safiee,
  • Muhammad Sharfi Najib,
  • Mujahid Mohamad,
  • Nazriyah Che Zan

摘要

Accurate classification of Palm Oil Fresh Fruit Bunch (POFFB) ripeness is crucial for optimizing yield and quality in the palm oil industry. Traditional manual methods are time-consuming and prone to error. This study presents an integrated approach utilizing Near-Infrared (NIR) optical sensors combined with artificial data generation and Case-Based Reasoning (CBR) to enhance classification accuracy and efficiency. NIR optical sensors capture spectral color data indicative of POFFB ripeness levels while artificial data generation is employed to provide comprehensive datasets overcoming the limitations and cost implications of preparing numerous real POFFB samples. The CBR method is used for classification and evaluated using performance metrics. The result shows that the classification model demonstrates a strong performance with an overall accuracy of 98.93%, sensitivity and precision both at 97.32%, and specificity at 99.33%. The study reveals the strong potential of these approaches to enhance efficiency and reliability in POFFB ripeness classification.