BibTex Citation Data :
@article{YPJ8412, author = {Putri Nuriskianti and Kusworo Adi and Tony Yulianto}, title = {KLASIFIKASI DAERAH LONGSOR BERBASIS PENGOLAHAN CITRA MENGGUNAKAN JARINGAN SYARAF TIRUAN PROPAGASI BALIK}, journal = {Youngster Physics Journal}, volume = {4}, number = {2}, year = {2015}, keywords = {Landslide index, color segmentation, artificial neural network.}, abstract = { Landslides are natural events that occur due to the ground movement of the earth's surface. The movement influenced by its constituent such as soil type, land use and intensity of rainfall in some place that causes a material such as ground was moving. Research on landslide done based on field surveys. The potential of a region in the category of landslides can be done by mapping parameters - parameters of landslides in the form of a calculation using the image of a network system that has been trained to predict the condition of an area. Image processing is done by segmenting color for any information presented in an image of landslides parameters. The color segmentation results performed labeling process to represent the information in the image. Then the landslides indices obtained from the manual calculation of weighting parameters. The result of the calculation is used as an instructional manual for the neural network. Where the value of the index 1 is the lowest level of landslide or safety category. While the index level 5 is the highest landslide or category of highly vulnerable to landslides. To process the data from the manual calculation in artificial neural network using backpropagation algorithm. The research data was training data and testing of tissue obtained from the manual calculation of weighting parameters landslides. Network training successfully conducted with a total accuration (index normal manual landslides and landslide index network) of 100% and accuration of test results 91,2% network. In the training data used 96 samples of data and test data as much as 34 data. Keywords: Landslide index, color segmentation, artificial neural network. }, issn = {2302-7371}, pages = {197--204} url = {https://ejournal3.undip.ac.id/index.php/bfd/article/view/8412} }
Refworks Citation Data :
Landslides are natural events that occur due to the ground movement of the earth's surface. The movement influenced by its constituent such as soil type, land use and intensity of rainfall in some place that causes a material such as ground was moving. Research on landslide done based on field surveys. The potential of a region in the category of landslides can be done by mapping parameters - parameters of landslides in the form of a calculation using the image of a network system that has been trained to predict the condition of an area.
Image processing is done by segmenting color for any information presented in an image of landslides parameters. The color segmentation results performed labeling process to represent the information in the image. Then the landslides indices obtained from the manual calculation of weighting parameters. The result of the calculation is used as an instructional manual for the neural network. Where the value of the index 1 is the lowest level of landslide or safety category. While the index level 5 is the highest landslide or category of highly vulnerable to landslides. To process the data from the manual calculation in artificial neural network using backpropagation algorithm.
The research data was training data and testing of tissue obtained from the manual calculation of weighting parameters landslides. Network training successfully conducted with a total accuration (index normal manual landslides and landslide index network) of 100% and accuration of test results 91,2% network. In the training data used 96 samples of data and test data as much as 34 data.
Keywords: Landslide index, color segmentation, artificial neural network.
Last update:
