Optimize Mosaic Dataset includes the following items:
Create Image Pyramid
Feature Description
Create an image pyramid for a mosaic dataset to improve the display efficiency of large volumes of image data. The application constructs a multilevel pyramid for the original images according to a defined rule, displaying pyramid images at corresponding resolutions at different scales. Moreover, when building overviews for the mosaic dataset, all images in the mosaic dataset must have pyramids to enhance data browsing speed.
Feature Entry
Image Panel->Other Data Sources->Specific Data Source Context Menu->Optimize Mosaic Dataset->Create Image Pyramid.
Parameter Description
- Resampling Method: The application provides the following resampling methods:
- Nearest, assigns the value of the nearest pixel in the input raster dataset to the corresponding pixel in the output raster dataset.
- Average, computes the mean of all valid values for resampling.
- Gaussian Kernel, uses a Gaussian kernel for resampling, which works well for images with high contrast and distinct pattern edges.
- Average Complex Data, computes the average in the magnitude-phase space for resampling images in complex data space.
- Bilinearity: assigns the weighted mean of the four nearest pixel values in the input raster dataset to the corresponding pixel in the output raster dataset. This method produces smoother results than Nearest but alters the original raster values.
- Cubic Convolution: similar to Bilinearity, assigns the weighted mean of the sixteen nearest pixel values in the input raster dataset to the corresponding pixel in the output raster dataset. This method yields the sharpest results and can enhance raster data edges, but it is computationally intensive and takes longer processing time.
- Cubic Linearity: based on the Akima interpolation algorithm, which, besides using the two measured points for interpolation between them, also uses the observations at the four neighboring points. In other words, Cubic Linearity considers six pixel values in the input raster dataset for resampling. This algorithm accounts for feature derivative values, resulting in a smooth and natural interpolation curve without unnatural oscillations.
- Lanczos Sine Resampling: uses a convolution filter by centering the convolution function at each resampling point and multiplying and summing all input values by the value of the convolution function at that position. The convolution function, also known as the Lanczos kernel, is based on sinc(x)=sin(x*pi)/x.
- Compression Method: The application provides three encoding types—DEFLATE, JPEG, and LZW—which allow users to balance pyramid display quality with storage space requirements according to practical needs. You can also select NONE to leave images uncompressed. For more details, see Dataset Compression Encode Type.
- Force Build Image Pyramid: Select this option to force the application to recreate and overwrite any existing image pyramids on the image.
Delete Image Pyramid
When importing image data with a large volume, image pyramids are usually created. However, after re-editing the image dataset, previously built image pyramids remain unchanged. In such cases, you must delete the existing image pyramids and rebuild them.
Feature Entry
Image Panel->Other Data Sources->Specific Data Source Context Menu->Optimize Mosaic Dataset->Delete Image Pyramid.
Create Histogram
Feature Description
Create a histogram for a mosaic dataset to reflect its data characteristics and determine the dataset’s stretch display effect. Histograms are created for all files in the mosaic dataset. If some images already have histograms and others do not, running Create Histogram will prompt a dialog saying, “Some images already have histograms. Do you want to recreate them?” Choose Yes to create histograms for all images, overwriting existing ones; choose No to create histograms only for images that do not already have them.
No interface is provided to view histograms.
Feature Entry
Image Panel->Other Data Sources->Specific Data Source Context Menu->Optimize Mosaic Dataset->Create Histogram.
Related Topics