Pipeline overview

Work in progress

Placeholder content. Being rewritten with figures.

Every ASP run, regardless of sensor, follows the same five-stage pattern.

The five stages

Sensor prep → bundle adjustment → orthorectification → stereo + DEM generation → alignment.

Stage 1: Sensor preparation

Vendor-specific tools (aster2asp, wv_correct, cam2map, dg_mosaic) convert raw vendor data into the standard image-plus-camera pair ASP expects.

Stage 2: Bundle adjustment

bundle_adjust refines the vendor camera models by minimizing reprojection errors of features matched between the input images. See Bundle adjustment.

Stage 3: Orthorectification

mapproject resamples each input image onto a reference DEM grid so stereo matching has only a small remaining disparity to solve. See Orthorectification.

Note

ASP calls this step “mapprojection” in its toolchain (mapproject, --mapproj-dem). This guide uses “orthorectification” — the more standard photogrammetry term — for the concept while keeping ASP’s tool names verbatim.

Stage 4: Stereo + DEM generation

parallel_stereo matches pixels between the two images and triangulates them into a point cloud; point2dem grids the cloud into a regular DEM. See Stereo photogrammetry.

Stage 5: Alignment

pc_align registers the DEM to a trusted reference (ICESat-2, MOLA, LOLA, or another DEM) using ICP. See Alignment.

What asp_plot does at every stage

Each ASP stage produces files; asp_plot reads them and produces diagnostic plots and PDF reports. See Visualization.