Cite Article

Classification of Jackfruit Fruit Damage Using Color Texture Features and Backpropagation Neural Network

Choose citation format

BibTeX

@article{IJASEIT8508,
   author = {Jonah Flor V. Oraño and Elmer A. Maravillas and Chris Jordan G. Aliac},
   title = {Classification of Jackfruit Fruit Damage Using Color Texture Features and Backpropagation Neural Network},
   journal = {International Journal on Advanced Science, Engineering and Information Technology},
   volume = {10},
   number = {5},
   year = {2020},
   pages = {1813--1820},
   keywords = {jackfruit; backpropagation neural network; color co-occurrence matrix; texture feature.},
   abstract = {

An accessible and cost-effective technology for plant pest and disease diagnosis could be beneficial for the farmers to be equipped with the technical know-how in producing high quality and quantity of crop yields. This study presents an implementation of image processing and machine learning techniques in building a predictive model for a computer-based and a mobile-based classification of jackfruit fruit damages caused by pests (fruit borer and fruit fly) and diseases (rhizopus fruit rot and sclerotium fruit rot). First, captured images of healthy, and infected fruit were split into two datasets: 60% for training and 40% for the testing phase, wherein each set contains five different classes. Then pre-processing methods such as cropping, scaling, and median filtering were applied that would make these images appropriate for information extraction. Next, 13 Haralick texture features were extracted from color co-occurrence matrices generated from Hue, Saturation, and Luminance color components of pre-processed images. Through Pearson’s correlation approach, texture features such as uniformity, variance, sum average, sum entropy, and entropy were selected as significant descriptors for training the classification model using a backpropagation learning algorithm. Lastly, basic evaluation metrics such as accuracy, precision, sensitivity, specificity, and Cohen’s kappa were computed to determine the performance of the model in recognizing the type of fruit damage on an unforeseen dataset. As a result, an overall accuracy rate of 93.42% and a kappa value of 0.9146 were obtained. In addition, the developed application displays suggestions on the proper pest control or disease management of the identified damage on the fruit surface of jackfruit.

},    issn = {2088-5334},    publisher = {INSIGHT - Indonesian Society for Knowledge and Human Development},    url = {http://ijaseit.insightsociety.org/index.php?option=com_content&view=article&id=9&Itemid=1&article_id=8508},    doi = {10.18517/ijaseit.10.5.8508} }

EndNote

%A Oraño, Jonah Flor V.
%A Maravillas, Elmer A.
%A Aliac, Chris Jordan G.
%D 2020
%T Classification of Jackfruit Fruit Damage Using Color Texture Features and Backpropagation Neural Network
%B 2020
%9 jackfruit; backpropagation neural network; color co-occurrence matrix; texture feature.
%! Classification of Jackfruit Fruit Damage Using Color Texture Features and Backpropagation Neural Network
%K jackfruit; backpropagation neural network; color co-occurrence matrix; texture feature.
%X 

An accessible and cost-effective technology for plant pest and disease diagnosis could be beneficial for the farmers to be equipped with the technical know-how in producing high quality and quantity of crop yields. This study presents an implementation of image processing and machine learning techniques in building a predictive model for a computer-based and a mobile-based classification of jackfruit fruit damages caused by pests (fruit borer and fruit fly) and diseases (rhizopus fruit rot and sclerotium fruit rot). First, captured images of healthy, and infected fruit were split into two datasets: 60% for training and 40% for the testing phase, wherein each set contains five different classes. Then pre-processing methods such as cropping, scaling, and median filtering were applied that would make these images appropriate for information extraction. Next, 13 Haralick texture features were extracted from color co-occurrence matrices generated from Hue, Saturation, and Luminance color components of pre-processed images. Through Pearson’s correlation approach, texture features such as uniformity, variance, sum average, sum entropy, and entropy were selected as significant descriptors for training the classification model using a backpropagation learning algorithm. Lastly, basic evaluation metrics such as accuracy, precision, sensitivity, specificity, and Cohen’s kappa were computed to determine the performance of the model in recognizing the type of fruit damage on an unforeseen dataset. As a result, an overall accuracy rate of 93.42% and a kappa value of 0.9146 were obtained. In addition, the developed application displays suggestions on the proper pest control or disease management of the identified damage on the fruit surface of jackfruit.

%U http://ijaseit.insightsociety.org/index.php?option=com_content&view=article&id=9&Itemid=1&article_id=8508 %R doi:10.18517/ijaseit.10.5.8508 %J International Journal on Advanced Science, Engineering and Information Technology %V 10 %N 5 %@ 2088-5334

IEEE

Jonah Flor V. Oraño,Elmer A. Maravillas and Chris Jordan G. Aliac,"Classification of Jackfruit Fruit Damage Using Color Texture Features and Backpropagation Neural Network," International Journal on Advanced Science, Engineering and Information Technology, vol. 10, no. 5, pp. 1813-1820, 2020. [Online]. Available: http://dx.doi.org/10.18517/ijaseit.10.5.8508.

RefMan/ProCite (RIS)

TY  - JOUR
AU  - Oraño, Jonah Flor V.
AU  - Maravillas, Elmer A.
AU  - Aliac, Chris Jordan G.
PY  - 2020
TI  - Classification of Jackfruit Fruit Damage Using Color Texture Features and Backpropagation Neural Network
JF  - International Journal on Advanced Science, Engineering and Information Technology; Vol. 10 (2020) No. 5
Y2  - 2020
SP  - 1813
EP  - 1820
SN  - 2088-5334
PB  - INSIGHT - Indonesian Society for Knowledge and Human Development
KW  - jackfruit; backpropagation neural network; color co-occurrence matrix; texture feature.
N2  - 

An accessible and cost-effective technology for plant pest and disease diagnosis could be beneficial for the farmers to be equipped with the technical know-how in producing high quality and quantity of crop yields. This study presents an implementation of image processing and machine learning techniques in building a predictive model for a computer-based and a mobile-based classification of jackfruit fruit damages caused by pests (fruit borer and fruit fly) and diseases (rhizopus fruit rot and sclerotium fruit rot). First, captured images of healthy, and infected fruit were split into two datasets: 60% for training and 40% for the testing phase, wherein each set contains five different classes. Then pre-processing methods such as cropping, scaling, and median filtering were applied that would make these images appropriate for information extraction. Next, 13 Haralick texture features were extracted from color co-occurrence matrices generated from Hue, Saturation, and Luminance color components of pre-processed images. Through Pearson’s correlation approach, texture features such as uniformity, variance, sum average, sum entropy, and entropy were selected as significant descriptors for training the classification model using a backpropagation learning algorithm. Lastly, basic evaluation metrics such as accuracy, precision, sensitivity, specificity, and Cohen’s kappa were computed to determine the performance of the model in recognizing the type of fruit damage on an unforeseen dataset. As a result, an overall accuracy rate of 93.42% and a kappa value of 0.9146 were obtained. In addition, the developed application displays suggestions on the proper pest control or disease management of the identified damage on the fruit surface of jackfruit.

UR - http://ijaseit.insightsociety.org/index.php?option=com_content&view=article&id=9&Itemid=1&article_id=8508 DO - 10.18517/ijaseit.10.5.8508

RefWorks

RT Journal Article
ID 8508
A1 Oraño, Jonah Flor V.
A1 Maravillas, Elmer A.
A1 Aliac, Chris Jordan G.
T1 Classification of Jackfruit Fruit Damage Using Color Texture Features and Backpropagation Neural Network
JF International Journal on Advanced Science, Engineering and Information Technology
VO 10
IS 5
YR 2020
SP 1813
OP 1820
SN 2088-5334
PB INSIGHT - Indonesian Society for Knowledge and Human Development
K1 jackfruit; backpropagation neural network; color co-occurrence matrix; texture feature.
AB 

An accessible and cost-effective technology for plant pest and disease diagnosis could be beneficial for the farmers to be equipped with the technical know-how in producing high quality and quantity of crop yields. This study presents an implementation of image processing and machine learning techniques in building a predictive model for a computer-based and a mobile-based classification of jackfruit fruit damages caused by pests (fruit borer and fruit fly) and diseases (rhizopus fruit rot and sclerotium fruit rot). First, captured images of healthy, and infected fruit were split into two datasets: 60% for training and 40% for the testing phase, wherein each set contains five different classes. Then pre-processing methods such as cropping, scaling, and median filtering were applied that would make these images appropriate for information extraction. Next, 13 Haralick texture features were extracted from color co-occurrence matrices generated from Hue, Saturation, and Luminance color components of pre-processed images. Through Pearson’s correlation approach, texture features such as uniformity, variance, sum average, sum entropy, and entropy were selected as significant descriptors for training the classification model using a backpropagation learning algorithm. Lastly, basic evaluation metrics such as accuracy, precision, sensitivity, specificity, and Cohen’s kappa were computed to determine the performance of the model in recognizing the type of fruit damage on an unforeseen dataset. As a result, an overall accuracy rate of 93.42% and a kappa value of 0.9146 were obtained. In addition, the developed application displays suggestions on the proper pest control or disease management of the identified damage on the fruit surface of jackfruit.

LK http://ijaseit.insightsociety.org/index.php?option=com_content&view=article&id=9&Itemid=1&article_id=8508 DO - 10.18517/ijaseit.10.5.8508