Mosaic Process

  V5.70 -- August 1997

The new version of the Mosaic process provides a new easy-to-use interface and additional mosaicking features. You can now combine raster images with different orientations and map projections, and select one of the input projections for the mosaicked raster. You can also include raster images of different data type in a single mosaic. You can mosaic different types of grayscale rasters, RGB raster sets with color composites, or even color with grayscale rasters. The input objects can be displayed in either full or wireframe mode in the View window. When the mosaic process is complete, the mosaic is automatically displayed in the Mosaic Results window.

There are a number of options that allow you to fine-tune the appearance of the mosaic. You can define the geographic extents of the mosaic manually by drawing an extents box, or match the extents to a reference object. The contrast matching option creates a mosaic with uniform contrast and color balance by matching histograms of the input objects to a histogram from one of the input objects (see below), from a reference object, or to a model normalized or equalized histogram. Histogram matching uses the whole extent of each input object, or just designated processing areas (see illustration to the right). Overlap areas can be processed in a number of different ways (including Average, Chessboard, and Random Feathering) to produce gradual, nearly invisible transitions between input images. When some corresponding input cells differ greatly in brightness (such as in images from different seasons), the Adaptive Filter can be used to override these overlap operations on a cell-by-cell basis and apply the Last Raster cell value for the problem cells instead.


Mosaic of five color-infrared 35 mm slides (scanned to RGB raster sets) covering adjacent mile-square sections, with one of the input slides shown to the right. The mosaic is displayed over a grayscale Digital Orthophoto Quad (DOQ) raster (which shows through the incomplete part of the mosaic in the upper left corner). The DOQ raster was used as a reference object to control the extents and output projection of the mosaic. A processing area was applied to each section-centered slide to mask the edges and place the seams at or near the section boundaries. Contrast matching to one of the input slides produced uniform color throughout the mosaic.

Draw a processing area for each input object to automatically mask unwanted parts of the image, such as fiducial marks, marginal data blocks, or the severely vignetted corners shown here. You can also use them to design custom seams between adjacent images.

The new Mosaic control window has process controls organized into several tabbed panels.


Histograms for red, green, and blue components are calculated automatically when you apply contrast matching to color composite images.