← Back to Skills Marketplace
wu-uk

egomotion-estimation

by wu-uk · GitHub ↗ · v0.1.0 · MIT-0
cross-platform ✓ Security Clean
76
Downloads
0
Stars
0
Active Installs
1
Versions
Install in OpenClaw
/install dynamic-object-aware-egomotion-egomotion-estimation
Description
Estimate camera motion with optical flow + affine/homography, allow multi-label per frame.
README (SKILL.md)

When to use

  • You need to classify camera motion (Stay/Dolly/Pan/Tilt/Roll) from video, allowing multiple labels on the same frame.

Workflow

  1. Feature tracking: goodFeaturesToTrack + calcOpticalFlowPyrLK; drop if too few points.
  2. Robust transform: estimateAffinePartial2D (or homography) with RANSAC to get tx, ty, rotation, scale.
  3. Thresholding (example values)
    • Translate threshold th_trans (px/frame), rotation (rad), scale delta (ratio).
    • Allow multiple labels: if scale and translate are both significant, emit Dolly + Pan; rotation independent for Roll.
  4. Temporal smoothing: windowed mode/median to reduce flicker.
  5. Interval compression: merge consecutive frames with identical label sets into start->end.

Decision sketch

labels=[]
for each frame i>0:
    lbl=[]
    if abs(scale-1)>th_scale: lbl.append("Dolly In" if scale>1 else "Dolly Out")
    if abs(rot)>th_rot: lbl.append("Roll Right" if rot>0 else "Roll Left")
    if abs(dx)>th_trans and abs(dx)>=abs(dy): lbl.append("Pan Left" if dx>0 else "Pan Right")
    if abs(dy)>th_trans and abs(dy)>abs(dx): lbl.append("Tilt Up" if dy>0 else "Tilt Down")
    if not lbl: lbl.append("Stay")
    labels.append(lbl)

Heuristic starting points (720p, high fps; scale with resolution/fps)

  • Tune thresholds based on resolution and frame rate (e.g., normalize translation by image width/height, rotation in degrees, scale as relative ratio).
  • Low texture/low light: increase feature count, use larger LK windows, and relax RANSAC settings.

Self-check

  • Fallback to identity transform on failure; never emit empty labels.
  • Direction conventions consistent (image right shift = camera pans left).
  • Multi-label allowed; no forced single label.
  • Compressed intervals cover all sampled frames; keys formatted correctly.
Usage Guidance
This skill is internally coherent and does not request credentials or perform network actions. Before installing: (1) understand that SKILL.md is a design/algorithm sketch, not runnable code — you (or the agent runtime) will need appropriate libraries (e.g., OpenCV, a Python runtime) to implement it; (2) confirm where and how video frames will be provided and whether any sensitive video would be processed or uploaded by your agent; (3) if you expect a packaged implementation, prefer a skill that declares exact dependencies or includes code/verifiable sources. Overall it appears fine from a security/permission perspective.
Capability Analysis
Type: OpenClaw Skill Name: dynamic-object-aware-egomotion-egomotion-estimation Version: 0.1.0 The skill bundle contains standard instructions and pseudocode for camera egomotion estimation using computer vision techniques (optical flow, RANSAC, and affine transforms). The SKILL.md file describes a legitimate workflow for classifying video motion (Pan, Tilt, Dolly, etc.) and lacks any indicators of malicious intent, data exfiltration, or harmful prompt injection.
Capability Assessment
Purpose & Capability
Name/description align with the instructions: feature tracking, RANSAC-based affine/homography estimation, thresholding and temporal compression. The only minor mismatch is that the SKILL.md does not declare required runtime dependencies (e.g., OpenCV, Python/runtime), which are necessary to implement the described functions but are not security-sensitive.
Instruction Scope
The instructions are narrowly scoped to processing video frames (feature detection, optical flow, transform estimation, label heuristics) and do not ask the agent to read unrelated files, access environment secrets, or transmit data to external endpoints.
Install Mechanism
No install spec (instruction-only), so nothing is written to disk or fetched during install — lowest install risk.
Credentials
The skill requests no environment variables, credentials, or config paths. That is proportionate to its stated local-video-processing purpose.
Persistence & Privilege
The skill is not forced-always and does not request elevated persistence or modify system/other-skill configs. Agent autonomous invocation is allowed (platform default) but not by itself a concern here.
How to Use
  1. Make sure OpenClaw is installed (local or Docker)
  2. Run the install command in chat: /install dynamic-object-aware-egomotion-egomotion-estimation
  3. After installation, invoke the skill by name or use /dynamic-object-aware-egomotion-egomotion-estimation
  4. Provide required inputs per the skill's parameter spec and get structured output
Version History
v0.1.0
Bulk publish from all-task-skills-dedup
Metadata
Slug dynamic-object-aware-egomotion-egomotion-estimation
Version 0.1.0
License MIT-0
All-time Installs 0
Active Installs 0
Total Versions 1
Frequently Asked Questions

What is egomotion-estimation?

Estimate camera motion with optical flow + affine/homography, allow multi-label per frame. It is an AI Agent Skill for Claude Code / OpenClaw, with 76 downloads so far.

How do I install egomotion-estimation?

Run "/install dynamic-object-aware-egomotion-egomotion-estimation" in the OpenClaw or Claude Code chat to install it in one step — no extra setup required.

Is egomotion-estimation free?

Yes, egomotion-estimation is completely free, licensed under MIT-0. You can download, install and use it at no cost.

Which platforms does egomotion-estimation support?

egomotion-estimation is cross-platform and runs anywhere OpenClaw / Claude Code is available (cross-platform).

Who created egomotion-estimation?

It is built and maintained by wu-uk (@wu-uk); the current version is v0.1.0.

💬 Comments