Researcher

Charles A. Rothwell

Also known as Charlie Rothwell

Charles A. Rothwell is a computer vision researcher whose work applied projective geometry to object recognition from images. He coauthored papers on invariant descriptors in 1990 and pose-independent 3-D recognition in 1991. His 1992 work addressed curved-surface outlines and pose recovery from plane curves, and his 1993 paper on projective structure won the ICCV Marr Prize. By 1995, he had left Oxford for INRIA’s Sophia Antipolis research unit, worked on the ROBOTVIS project and developed projective indexes in his book Object Recognition through Invariant Indexing.

Timeline

  1. 1990

    Relative motion and pose

    At the British Machine Vision Conference in 1990, he coauthors ‘Relative Motion and Pose from Invariants.’ Its projectively invariant shape descriptors identify object models in images without reference to object pose.

  2. 1991

    Invariants for 3-D objects

    In 1991, he coauthors ‘Invariant Descriptors for 3-D Object Recognition and Pose’ with five collaborators. Their model-based system recognizes curved plane objects regardless of pose and describes invariant features for 3-D objects with plane faces.

  3. 1992

    Outlines of curved surfaces

    In 1992, he coauthors ‘Recognising Rotationally Symmetric Surfaces from their Outlines’ for ECCV. The method uses bitangencies, creases and inflections in image outlines to describe curved surfaces across viewpoints.

  4. 1992

    Pose from plane curves

    In 1992, he coauthors ‘Relative Motion and Pose from Arbitrary Plane Curves.’ The paper recovers the position and orientation of known coplanar conics from a single perspective image and extends pose methods to arbitrary plane curves.

  5. 1993

    The ICCV Marr Prize

    In 1993, his paper ‘Extracting Projective Structure from Single Perspective Views of 3D Point Sets’ wins the ICCV Marr Prize.

  6. 1994

    The MORSE system

    At the ARPA Image Understanding Workshop in 1994, he coauthors the MORSE paper on 3-D object recognition. The system uses geometric invariants from one 2-D intensity view and applies object-class constraints to feature extraction and grouping.

  7. 1994

    Driving vision by topology

    In 1994, he coauthors INRIA Technical Report 2444, ‘Driving Vision by Topology.’ It argues that perceptual grouping works more efficiently and reliably when well-formed topological structures are available.

  8. By 1995

    INRIA and invariant indexing

    By 1995, he leaves Oxford for INRIA’s research unit at Sophia Antipolis and works on the ROBOTVIS project. His 1995 book, ‘Object Recognition through Invariant Indexing,’ develops projective indexes for a recognition system.

Career

  1. Until 1995University of Oxford
  2. From 1995Inria

Sources and credits

Photo: We have no published photo of Charles A. Rothwell yet. Send us a free-licensed photo.

  1. Visual Geometry Group - University of Oxford robots.ox.ac.uk. Checked 21 October 2026.
  2. Invariant Descriptors for 3-D Object Recognition and Pose - Illinois Experts experts.illinois.edu. Checked 21 October 2026.
  3. Recognising rotationally symmetric surfaces from their outlines - Illinois Experts experts.illinois.edu. Checked 21 October 2026.
  4. Visual Geometry Group - University of Oxford robots.ox.ac.uk. Checked 21 October 2026.
  5. Visual Geometry Group - University of Oxford robots.ox.ac.uk. Checked 21 October 2026.
  6. Visual Geometry Group - University of Oxford robots.ox.ac.uk. Checked 21 October 2026.
  7. Papers available from krakatoa.inria.fr robots.ox.ac.uk. Checked 21 October 2026.
  8. CiNii 図書 - Object recognition through invariant indexing ci.nii.ac.jp. Checked 21 October 2026.
  9. Charlie ROTHWELL robots.ox.ac.uk. Checked 21 October 2026.