A sample of the AutoCNet library being used for pseudo-dense matching on straight-forward Apollo 15 metric camera images. Blue points are identified correspondences and yellow lines indicate that the points are linked (correspond with one another).

AutoCNet

AutoCNet is an open source library (API) for the pseudo-automated generation of sparse n-image control networks. AutoCNet is licensed in the public domain.

AutoCNet supports the following:

• 32-bit CUDA based (GPU) keypoint extraction

• Support for images of arbitrary size using downsampling or tiling for feature extraction

• A reference implementation (the CandidateGraph) to demonstrate use of the API

Tutorials

An example is worth 1000 pages of documentation. Checkout our tutorials (jupyter notebooks).

• Installation
• Working With ISIS Control Networks
• Working With Apollo Pan
• Change Detection
• Astrogeology Science Center Workshop

Issues and Version Control

This project is hosted on Github. If you run into a problem, please open an issue on our issue tracker.

Documentation

• Library Reference
  • camera — CV Camera Functions
    • camera.camera — CV Camera Functions
    • camera.utils — CV Utility Functions
  • cg — Computational Geometry
    • cg.cg — Computational Geometry
    • cg.change_detection — Change Detection Algorithms
  • control — Control
    • control.control — Control Network
  • graph — Graph Based Data Representaton
    • graph.network — Network File Representation
    • graph.node — Network Node Object
    • graph.edge — Network Edge
    • graph.markov_cluster — Markov Clustering Algorithm
  • io — File Input/Output
    • io.keypoints — Keypoint Serialization
    • io.network — Network Serialization
    • io.db — Database Models for AutoCNet Server
      • io.db.connection — Database Connection
      • io.db.controlnetwork — Database to Control Networks
      • io.db.model — Database Models
      • io.db.redis_queue — Redis Queue Management
      • io.db.triggers — Database Triggers
      • io.db.wrappers — Database Wrappers to Data Structures
  • matcher — Graph Based Data Representaton
    • CPU Based Feature Extraction and Matching
      • matcher.cpu_decompose — CPU Based Recursive Image Decomposition
      • matcher.cpu_extractor — Extracting Features from Images
      • matcher.cpu_matcher — CPU Based n-Keypoint Matcher
      • matcher.cpu_outlier_detector — Finding outliers in both the keypoints and the matches
      • matcher.cpu_ring_matcher — CPU Based Ring Matching
    • CUDA Based Feature Extraction and Matching
      • matcher.cuda_extractor — GPU Based n-Feature Extractor
      • matcher.cuda_decompose — GPU Based Recursive Image Decomposition
      • matcher.cuda_matcher — GPU Based n-Keypoint Matcher
      • matcher.cuda_outlier_detector — Finding outliers in both the keypoints and the matches
    • Subpixel Refinement
      • matcher.subpixel — Performing subpixel registration between images
      • matcher.ciratefi — RST-BC Invariant template match algorithm
      • matcher.naive_template — Simple template matching
    • High-Level Matchiing Examples
      • matcher.cross_instrument_matcher — Multi-imstrument Matching Methods
    • Spatial Suppression Function Examples
      • matcher.suppression_funcs — Pre-Generated Generic Suppression Functions
  • spatial — Spatial
    • spatial.isis — Wrapper to ISIS Sensors
    • spatial.overlap — Point Placement in Overlaps
  • transformation — Transformation Objects
    • Reprojection Transformations
      • transformation.homography — Homographies
      • transformation.fundamental_matrix — Fundamental Matrices
      • transformation.spatial — Spatial
    • Image Decomposition
      • transformation.decompose — Coupled Decomposition
  • utils — Utility Functions
    • utils.utils — Utility Functions
  • vis — Data Visualization
    • vis.graph_view — Graph Visualization

Developers

• Contributing
  • Types of Contributions
  • Get Started!
  • Pull Request Guidelines
  • Tips
• Credits
  • Development Team
  • Contributors

References

• References

License

• Unlicense