top: paper
url: https://pdfs.semanticscholar.org/e75b/cc88fba365798e31e0d607d173f71c3af7c4.pdf

note

• how they came up with fluency (effective of the movement) and agility
  • that custom definition might be a bit outdated now considering that research got published in 200x
    • what is the latest approach?
• biomechanic?
• (iROZHLAS, 2019) how they calculate and came up with that conclusion about the climber’s physique advantage
• PCA and bouldering
• K-shortest path and motion planning
• XR approaches

ideas

• Collaboration climbing ? (inspired by it takes two and teaching/training)

note (jot down some keywords)

intro

• climbing motion analysis
  • to optimize climbing techniques
  • therapeutic application: avoid movements that are prone to cause injuries
• sensors
• human motion capture
• analysis algorithm
• RGB-D camera
• section 2: sensor techs
• section 3: review motion capture approaches
• section 4: algorithms for climbing motion analysis
• section 5: the findings

sensors

• optical sensors
  • pros: contactless
• camera technology with depth sensing
• RGB-D camera
• triangulation process — depth estimation
• Time of flight (ToF)
• force sensors

motion capture

• Human pose estimation (HPE)
  • input: locations of defined points (body)
    • coarse body description
      • center of mass (CoM)
    • fine-granular models
      • skeleton model
    • 2D vs 3D representation
• CoM
  • desired output:
    • entropy
    • velocity
    • acceleration
  • translated to:
    • fluency
    • force of the motion (2007)
  • how to calculate
    • weighted average of nine body segments
    • 3-D skeleton
    • 3-D point clouds

Pose Estimation for Climbing Analysis

• mostly rely on fine granular skeleton model
• instrumented bouldering wall
  • force torque sensor
• A very popular Czech competitive climber, Adam Ondra, ”hung with sensors” to analyse what makes him such an outstanding climber (iROZHLAS, 2019)
  • marker-based motion capture system
  • result: Next to his long neck, his comparatively slim shoulders result in less force on his fingers.
• parsing a climber’s body area
  • foreground segmentation on depth image
    • used to correct the feet and hand positions of the kinetic skeleton
  • seems to related with these researches
    • 2017
      • https://www.researchgate.net/publication/319470338_A_climbing_motion_recognition_method_using_anatomical_information_for_screen_climbing_games
    • https://liu.diva-portal.org/smash/get/diva2:2028608/FULLTEXT01.pdf
• ML based pose estimation
  • problem (2018)
    • finding specific dataset for climbing pose
• more keywords
  • high degree of freedom
  • self-occlusion

Algorithms for motion analysis

• PCA 2007
• MoCap system with passive sensors (Sibella et al., 2007
  • CoM -> entropy, velocity, acceleration
  • fluency: effectiveness of the movement
    • means of the entropy of the climbing route
  • agility
    • speed and acceleration
• instrumented climbing wall
  • 3-D camera
  • position and climber’s force
  • motion sensor -> hip
  • dynos
    • relation between vertical and absolute velocities of the hip
• motion planning
  • prediction/plan climbing motion
  • anatomically possible movements
  • data driven analysis
  • simulating climbing behavior
    • let climbers design quality routes by placing holds in a virtual climbing wall
  • offline route planning
    • graph based application
    • k-shortest path (2017)
    • depends on the anatomic characteristic of the climber
      • strength
      • flexibility
      • reach
      • contemplated limbs hang free for balancing and the use of the wall friction
    • simulated agent can move more than one limb at a time
  • AR
    • create climbing routes for moon board with their smartphone
  • VR
    • share the experiences professional climbers made on extreme routes
    • Adidas, 2019
• Teaching and Training
  • demonstration of postures and movements / imitate
  • biomechanics point of view
  • visualizing reference motion
    • difficulty to teach and learn simultaneously
    • augmented real-time video projected on the wall

summary and outlook

• at 2018, real-time marker-less, vision-based motion capture for climbing motions is far from being solved and requires further research activities.

created on: Tue Feb 17 2026