Stage of maturity banana fruit classification using image processing / Nadia Kasim ... [et al.]

• Main Authors: Kasim, Nadia, Zainol Abidin, Siti Nazifah, Abu Mangshor, Nur Nabilah, Mohamed Hamzah, Hazwa Hanim, Abd Rahim, Nurul Zahirah
• Format: Article
• Language: English
• Published: Universiti Teknologi MARA, Perak 2019
• Subjects: Analytical methods used in the solution of physical problems; Algorithms
• Online Access: http://ir.uitm.edu.my/id/eprint/39632/1/39632.pdf; http://ir.uitm.edu.my/id/eprint/39632/; https://mijournal.wixsite.com/index
• Institution: Universiti Teknologi Mara

Maturity stage of fresh banana fruit is a crucial aspect of import and export industry. The ability to identify maturity stage of banana will be a great support but only experts can determine the maturity stage accurately. Lack of knowledge and skill of maintaining fruit quality standard to match the international standard is a problem faced by the farmers. This study attempted to propose a system that uses image processing to detect the maturity stage of banana based on its color and size using Support Vector Machine (SVM) learning algorithm. Human errors and cost of hired expert in detecting maturity of fresh banana fruit can be reduced. A total of 45 images are tested and only three stages are considered in this study which are stage one, stage two and stage six out of seven stages. Stage one and stage two is the stage suitable for the international standard, while stage six is for a national standard of import export activities. The system showed 88.89% accuracy classification by comparing the result yield by the system with results offered by the experts. Hence, the proposed system can be commercially used for export and import activities as it can automatically determine the maturity stage of fresh banana and help the farmers in making the right decision on harvesting.