Hyperspectral Image Mosaicking Based on Double-Layer Fusion of Image and Data

The traditional mosaicking technology exhibits insufficient utilization of the image information. Therefore, a hyperspectral image mosaicking technique based on the double-layer fusion of image and data is proposed. In case of the image layer, the scale-invariant feature transformation algorithm is used to extract the image features and the Euclidean distance is used to determine the feature matching range. Further, the features are matched according to the coordinate conversion relation to complete image layer mosaicking. In case of the data layer, the data is divided into high and low data. Then, the weighted sum method is used to calculate the new value of data and stitch it, and the high and low data are merged via the displacement operation to complete the mosaicking of the data layer. Finally, the image and data are stored in the BIL mode for completing the double-layer fusion of image with data. The hyperspectral image mosaicking experiment is conducted in a certain area. Experimental results demonstrate that the average mosaicking accuracies of the image and data layers are 0.9214 and 0.9663, respectively, indicating the effectiveness and accuracy of the proposed technique.