Tools
Tools
Premakar, Monica, Doss, Srinath, Virdee, Bal Singh and Khanna, Ashish (2026) DRIN with Jaya Flamingo Search Optimization for crop identification using hyperspectral satellite image. Iranian Journal of Science and Technology. pp. 1-57. ISSN 2364-1827
Crops are fundamental to human survival, and their importance rises as the global population continues to increase. The conventional methods for crop identification from hyperspectral images suffer from low accuracy issues and fail to effectively handle the high-dimensional data and redundant information. Thus, a novel methodology called Deep Residual Involution Network Jaya Flamingo Search Optimization (DRIN_JayaFSO) is presented for crop identification using Hyperspectral Satellite images. Primarily, the input hyperspectral image is forwarded to the Wiener filter in the preprocessing phase to eliminate the naturally present noise in the hyperspectral image. Further, the 3D Convolutional Autoencoder (3D-CAE) is utilized for image segmentation, and it is tuned using the Chronological Average Subtraction Based Optimization (CASBO). Moreover, the preprocessed hyperspectral image is passed into the feature extraction phase for the mining of the vegetative index and spatial features. Finally, the excerpted features are fed into the classification process, wherein the hyperspectral image is classified using the DRIN structurally optimized by the designed Jaya Flamingo Search Optimization (JayaFSO). The JayaFSO is established by merging the Jaya Optimization (Jaya) Algorithm and the Flamingo Search Algorithm (FSA). Further, DRIN_JayaFSO achieved superior accuracy, TPR, and TNR of 94.766%, 95.268%, and 91.877%.
![[thumbnail of Revised Manuscript - accepted 22-08-2026.pdf]](https://repository.londonmet.ac.uk/style/images/fileicons/text.png "Revised Manuscript - accepted 22-08-2026.pdf")
Restricted to Repository staff only until 23 September 2027.
Available under License Creative Commons Attribution Non-commercial No Derivatives 4.0.
Download (1MB) | Request a copy
 |
View Item |