Universiti Teknologi Malaysia Institutional Repository

Image analysis techniques for ripeness detection of palm oil fresh fruit bunches

Ghazalli, Shuwaibatul Aslamiah and Selamat, Hazlina and Omar, Zaid and Yusof, Rubiyah (2019) Image analysis techniques for ripeness detection of palm oil fresh fruit bunches. Journal of Electrical Engineering, 18 (3). pp. 57-62. ISSN 0128-4428


Official URL: https://dx.doi.org/10.11113/elektrika.v18n3.192


Being one of the biggest producers and exporters of palm oil and palm oil products, Malaysia has an important role to play in fulfilling the growing global need for oils and fats sustainably. Quality is an important factor that ensuring palm oil industries fulfill the demands of palm oil product. There has significant relationship between the quality of the palm oil fruits and the content of its oil. Ripe FFB gives more oil content, while unripe FFB give the least content. Overripe FFB shows that the content of oil is deteriorates. There have 4 classes of ripeness stages involves in this paper which are ripe, unripe, underipe and overripe. The proposed approach in this paper uses color features and bag of visual word for classifying oil palm fruit ripeness stages. Experiments conducted in this paper consisted of smartphone camera for image acquisition, python and matlab software for image pre processing and Support Vector Machine for classification. A total of 400 images is taken in a few plant in north Malaysia. Experiments involved on a dataset of 360 images for training for four classes and 40 images for testing. The average accuracy for the 4 classes of the FFB by color features is 57% while the accuracy for ripeness classification by using bag of visual word is 70%.

Item Type:Article
Uncontrolled Keywords:image processing , classification , bag of visualword, SVM
Subjects:T Technology > TK Electrical engineering. Electronics Nuclear engineering
Divisions:Electrical Engineering
ID Code:85276
Deposited By: Fazli Masari
Deposited On:17 Mar 2020 08:10
Last Modified:17 Mar 2020 08:10

Repository Staff Only: item control page