|
Scientific Visualization
| Issue Year: | 2017 |
| Quarter: | 1 |
| Volume: | 9 |
| Number: | 1 |
| Pages: | 73 - 84 |
|
| Article Name: |
ON SOME ASPECTS OF INTEGRATED MULTISPECTRAL SINGLE-LAYER IMAGING ARRAYS AND THEIR APPLICATION FOR IMAGE FUSION |
| Authors: |
M. Smagin (Russian Federation) |
| Address: |
M. Smagin
smagin@mniti.ru
Joint-Stock Company MNITIá Russian Federation |
| Abstract: |
Modern advancements in electronics and physics have created the opportunity to develop new types of multispectral image sensors with single or multiple sensitive layers. The distinctive feature of single-layer imaging arrays is the arrangement of its elements (sensible in different spectral bands) in a single plane that form a united array organized as a mosaic. Practical application of such imaging arrays for image fusion allows us to gain some important advantages, but it also requires the modification of fusion algorithms by adding the demosaic operation into the sequence of operations. It also creates the need to determine the proper spatial arrangement of sensible elements in the single-layer array, which allows the achievement of the best results, i.e. the most informative fused images.
The following article describes the result of a case study focused on the task to determine the robustness of several well-known image fusion algorithms (PCA, DWT and Laplacian Pyramid) to mosaic images and different types of spatial arrangements of sensors in multispectral image acquisition arrays. Several couples of source images (obtained in different spectral bands) were used to test the performance of the abovementioned algorithms on mosaicked images in order to find the most effective position of the demosaic operation in the sequence of operations in the tested algorithms. |
| Language: |
English |
|
|
|
