Lesson 1DSM vs DTM production: methods to create bare-earth DTM in partially vegetated, built areas and river channel modelingThis section explains DSM and DTM concepts and workflows, focusing on bare-earth extraction in vegetated and urban areas. Learners will apply filtering, editing, and river channel modeling to obtain hydrologically consistent terrain surfaces for local use.
DSM versus DTM definitions and usesGround point filtering in vegetationHandling buildings and urban artifactsManual terrain editing and breaklinesRiver channel and floodplain modelingLesson 2Orthomosaic generation: seamlines, radiometric blending, color balancing, and handling water/reflective surfacesThis section explains orthomosaic generation from DSM or DTM, including seamline design, radiometric balancing, and color correction. Learners will handle water, shadows, and reflective surfaces to produce visually consistent, accurate mosaics in regional contexts.
Orthorectification using DSM or DTMSeamline placement and editingRadiometric balancing and blendingColor correction and tone mappingHandling water, shadows, and glareLesson 3Bundle adjustment: theory overview, role in defining relative geometry, interpreting residuals and report metricsThis section introduces bundle adjustment theory and its role in defining the relative camera geometry. Learners will understand normal equations, constraints, and how to interpret residuals, covariance, and software reports to judge solution stability for South Sudan projects.
Collinearity equations and observation modelWeighted least squares and constraintsRelative orientation and network geometryInterpreting residuals and outliersReading bundle adjustment quality reportsLesson 4Image alignment and keypoint matching: objectives, tie points, and strategies to improve tie qualityThis section details image alignment and keypoint matching, including feature detection, tie point generation, and quality improvement strategies. Learners will tune alignment settings and diagnose poor overlap, blur, or repetitive texture issues in local imagery.
Feature detection and descriptor choiceTie point matching and cleaningAlignment settings and accuracy levelsDealing with low overlap or motion blurManaging water, snow, and textureless areasLesson 5Image import, metadata validation, and pre-processing: EXIF checks, lens profile, radiometric correctionsThis section covers safe image import, metadata validation, and pre-processing. Learners will verify EXIF integrity, lens profiles, and coordinate tags, then apply radiometric corrections and image culling to ensure robust downstream processing for regional data.
Organizing image datasets and backupsEXIF and GNSS metadata validationLens profile selection and verificationRadiometric and color correctionsImage quality screening and cullingLesson 6Vector extraction and planimetric mapping: breakline creation, building footprint extraction, road centerlines, and hydrographyThis section covers vector extraction and planimetric mapping from orthomosaics and elevation data. Learners will digitize breaklines, building footprints, roads, and hydrography, and apply topology rules and accuracy checks for mapping products in South Sudan.
Digitizing and snapping best practicesBreakline creation from terrain modelsBuilding footprint and roofline mappingRoad centerlines and pavement edgesHydrography and drainage network tracingLesson 7Camera calibration and interior orientation: self-calibration parameters, when to fix vs estimate focal length and distortionThis section covers camera calibration and interior orientation, including self-calibration parameters and when to fix or estimate them. Learners will understand focal length, principal point, and distortion models and their impact on accuracy in local surveys.
Interior orientation and camera modelsRadial and tangential distortion termsSelf-calibration requirements and risksFixing versus estimating focal lengthUsing pre-calibrated metric camerasLesson 8Dense point cloud generation: filtering, noise removal, classifying points (ground, vegetation, buildings)This section focuses on dense point cloud generation from aligned images. Learners will configure depth map settings, manage noise, classify points into ground and objects, and prepare a clean point cloud for surface modeling and editing in South Sudanese terrains.
Depth map generation parametersPoint density versus processing timeNoise filtering and outlier removalGround and non-ground classificationExport formats and data managementLesson 9Quality checks at each stage: tie point residuals, GCP residuals, RMSE on check points, visual comparison, cross-sections and profile checksThis section explains how to evaluate quality at each processing stage using quantitative metrics and visual checks. Learners will interpret residuals, RMSE, and profiles to detect issues early and decide when reruns or edits are required for reliable outputs.
Tie point residuals and reprojection errorGCP and check point residual analysisRMSE computation and interpretationVisual inspection of orthos and 3D viewCross-sections and elevation profile checksLesson 10Processing strategies for limited budgets: cloud vs local processing, open-source (e.g., OpenDroneMap) vs commercial software trade-offsThis section compares processing strategies under limited budgets, covering hardware choices, cloud versus local workflows, and open-source versus commercial tools. Learners will balance cost, speed, scalability, and data security for real projects in South Sudan.
Estimating compute and storage needsCloud processing costs and constraintsLocal workstation and GPU considerationsOpen-source photogrammetry toolchainsCommercial software features and licensing
