Iranian Research Organization for Science and Technology (IROST) Innovative Food Technologies 2783-350X 8 1 2020 10 22 Design and construction of an automatic detection system for orange defects using an attunable lightness algorithm Design and construction of an automatic detection system for orange defects using an attunable lightness algorithm 13 28 917 10.22104/jift.2020.3518.1844 FA Hadis Biabi Mechanics of Biosystems Engineering Department, Faculty of Agricultural Engineering and Rural Development, Agricultural Sciences and Natural Resources University of Khuzestan. Saman Abdanan Mehdizadeh Assistant professor of Khuzestan Agricultural Sciences and Natural Resources University Journal Article 2019 03 01 The automatic detection of defective fruit through the computer vision system continues to be a problem due to the uneven instability distribution on the citrus surface. As a result of the development of a system that is capable of detecting damage in citrus with high accuracy and speed is essential. Therefore, an adaptive lightness correction algorithm was implemented in this paper that simply overcomes the disturbance of the indirect distribution intensity in the fruit level in online and static conditions and avoids error detection. In the study, 200 specimens containing 50 healthy oranges and 150 defective oranges (Green Fruit Molds, Diaspididae, Alternaria Fruit and mechanical damage) were investigated. In this system, 4 images were taken from each sample and after applying the proposed algorithm, all four oranges were categorized into healthy and defective groups. Based on the results, it was found that the accuracy of the system for the damage of Green Fruit Molds, Diaspididae, Alternaria Fruit and mechanical damage was 87.80, 71.42, 74.28 and 100, indicating high performance of the proposed method. The automatic detection of defective fruit through the computer vision system continues to be a problem due to the uneven instability distribution on the citrus surface. As a result of the development of a system that is capable of detecting damage in citrus with high accuracy and speed is essential. Therefore, an adaptive lightness correction algorithm was implemented in this paper that simply overcomes the disturbance of the indirect distribution intensity in the fruit level in online and static conditions and avoids error detection. In the study, 200 specimens containing 50 healthy oranges and 150 defective oranges (Green Fruit Molds, Diaspididae, Alternaria Fruit and mechanical damage) were investigated. In this system, 4 images were taken from each sample and after applying the proposed algorithm, all four oranges were categorized into healthy and defective groups. Based on the results, it was found that the accuracy of the system for the damage of Green Fruit Molds, Diaspididae, Alternaria Fruit and mechanical damage was 87.80, 71.42, 74.28 and 100, indicating high performance of the proposed method. https://jift.irost.ir/article_917_da0d1111d2dc5d489242e60ebcbaf988.pdf