Intelligent mapping of irrigated areas from Landsat 8 images using transfer learning

Authors : Zouhair BENBAHRIA, İmane SEBARİ, Hicham HAJJİ, Mohamed Faouzi SMİEJ

Pages : 40-50

Doi:10.26833/ijeg.681312

View : 8 | Download : 4

Publication Date : 2021-02-01

Article Type : Research Paper

Abstract :The lack of reliable and up-to-date data in developing countries is a major obstacle to sustainable development. In Morocco, where groundwater withdrawals by farmers are very intensive and informal, maps describing and monitoring the extension of irrigated areas are scarce and labor-intensive to obtain. In this paper a novel transfer learning algorithm is proposed to map irrigated areas at different stages of an agricultural cycle from Landsat 8 images. The results obtained displays satisfactory performance over traditional machine learning algorithms. On a small dataset, we initially tested three well known deep learning architectures insert ignore into journalissuearticles values(SegNet, DenseNet and Unet);. The results obtained were not satisfactory. So, to get high performance, we rely on a transfer learning architecture combining UNet with ResNet50 backbone insert ignore into journalissuearticles values(trained on 2012 ILSVRC ImageNet dataset); as a baseline after a phase where different configurations were tested. In the first part of this study, we compared the use of three optimization methods: Adam and two variants of Stochastic Gradient Descent insert ignore into journalissuearticles values(SGD); associated with two techniques insert ignore into journalissuearticles values(Cyclical Learning Rate and Warm Restart); to find the optimal learning rate and then test the impact of data augmentation on the overall accuracies. Data augmentation had improved the overall accuracy for the three methods. Adam based method from 94% to 97% with mean IoU of 0,79 insert ignore into journalissuearticles values(for all land cover classes); and 0,86 for irrigated areas class. For SGD based methods, the overall accuracy had increased from 91% to 94% with mean IoU of 0,75 insert ignore into journalissuearticles values(for all land cover classes); and 0,82 for irrigated areas class. As we are interested in having irrigated areas maps at different key periods of the agricultural cycle, we also explored, in the second part of this study, the temporal generalization of the best model. 
Keywords : Irrigated areas mapping, Landsat 8, Deep learning, UNet, ResNet50