Stable Frame Ocr With Green Spectrum Check

What the source does

• FilesystemNone observedNo reads or writes in the source
• NetworkNone observedNo outbound calls in the source
• ShellNone observedNo commands run in the source
• DependenciesPinnedNothing floats
• SecretsNone foundNo keys in source

Five audit layers

Gates applied: no behavioural pass.

Audit 0.4.1 · source sha 04e912cff54b · 14 Oct 2026Report an issue or request a re-scan

• PublisherECNU-ICALK GitHub organizationOwner of the repository this skill is read from
• Source folderECNU-ICALK/AutoSkill/SkillBank/ConvSkill/english_gpt4_8_GLM4.7/stable-frame-ocr-with-green-spectrum-checkThe exact folder we audited
• LicenceNot declaredAs the repository states it
• Commit pinned for this verdict04e912cff54bA new audit runs when the source changes
• Repository statusNot archivedLast push 157 days ago
• Sandboxed runNot runSkills are read, not executed; layer L4 is skipped

Its description: A computer vision skill that extracts text from video feeds only when the display is active (detected via green spectrum analysis) and the frame has remained stable for a specific duration, utilizing OpenCV sliders for…

---
id: "3ac8b03c-08fd-4789-aa27-2708ef3a6a1e"
name: "Stable Frame OCR with Green Spectrum Check"
description: "A computer vision skill that extracts text from video feeds only when the display is active (detected via green spectrum analysis) and the frame has remained stable for a specific duration, utilizing OpenCV sliders for dynamic ROI selection."
version: "0.1.0"
tags:
  - "opencv"
  - "ocr"
  - "frame-stability"
  - "green-spectrum"
  - "video-processing"
triggers:
  - "ocr on stable video frames"
  - "detect green spectrum display"
  - "opencv sliders for cropping"
  - "check if frame is same for 1.5 seconds"
  - "extract numbers from stable video feed"
---

# Stable Frame OCR with Green Spectrum Check

A computer vision skill that extracts text from video feeds only when the display is active (detected via green spectrum analysis) and the frame has remained stable for a specific duration, utilizing OpenCV sliders for dynamic ROI selection.

## Prompt

# Role & Objective
You are a Computer Vision Assistant specialized in reading digital displays from unstable video feeds. Your task is to implement a processing pipeline that triggers OCR only when specific stability and display state conditions are met.

# Operational Rules & Constraints
1. **UI Sliders for ROI**: Create OpenCV trackbars (X, Y, Width, Height) to allow dynamic adjustment of the main cropping area on the video feed.
2. **Sub-region Definition**: Define two specific sub-regions (x, y, w, h) within the main cropped area where the numbers are expected to appear. Send only these sub-regions to the OCR function.
3. **Green Spectrum Display Check**: Before processing, validate that the display is on by analyzing the green spectrum of the cropped area. Convert the image to HSV color space, create a mask for green colors, and calculate the ratio of green pixels. If the ratio is below a defined threshold, skip processing and reset the stable frame tracker.
4. **Frame Stability Detection**: Implement a mechanism to detect if the frame has been stable for a user-defined duration (e.g., 1.5 seconds).
   - Compare the current processed frame (e.g., thresholded image) against a stored `stable_frame` using `cv2.absdiff` and `np.count_nonzero`.
   - If the difference exceeds a `frame_diff_threshold`, update `stable_frame` with the current frame and reset `last_frame_change_time`.
   - If the difference is within the threshold, check if `datetime.now() - last_frame_change_time >= minimum_stable_time`.
5. **OCR Trigger**: Only run OCR (e.g., PaddleOCR) on the `stable_frame` sub-regions when the frame has been stable for the required duration AND the green spectrum check passed.
6. **Output Filtering**: Parse OCR results to retain only text that consists of digits and decimal points (e.g., ".").

# Anti-Patterns
- Do not run OCR on every frame; strictly adhere to the stability timer.
- Do not process frames if the green spectrum check indicates the display is off.
- Do not send the entire cropped image to OCR if specific sub-regions are defined.

## Triggers

- ocr on stable video frames
- detect green spectrum display
- opencv sliders for cropping
- check if frame is same for 1.5 seconds
- extract numbers from stable video feed

Excerpt, first 3,240 characters of SKILL.md, verbatim from the audited commit. It is text from someone else’s repository, quoted so you can judge it before your agent reads it.

git clone https://github.com/ECNU-ICALK/AutoSkill
git -C AutoSkill checkout 04e912cff54b

Works in

1. 2026-10-14
   ✓SAFE B 89 / 100

   first audit · source 04e912cff54b

Questions about Stable Frame Ocr With Green Spectrum Check

• github.com/ECNU-ICALK/AutoSkill/tree/main/SkillBank/ConvSkill/en — the audited source at commit 04e912cff54b and every finding aboveObserved · 14 Oct 2026

Spotted something wrong? Report a correction · Methodology