gemini-logo-remover
π―Skillfrom bear2u/my-skills
Removes Gemini logos, watermarks, and AI-generated image markers using precise OpenCV inpainting techniques.
Installation
npx skills add https://github.com/bear2u/my-skills --skill gemini-logo-removerSkill Details
Remove Gemini logos, watermarks, or AI-generated image markers using OpenCV inpainting. Use this skill when the user asks to remove Gemini logo, AI watermark, or any logo/watermark from images.
Overview
# Gemini Logo Remover
Remove Gemini logos and watermarks from AI-generated images using inpainting.
Setup
```bash
pip install opencv-python numpy pillow --break-system-packages
```
Usage
By Coordinates
```python
import cv2
import numpy as np
def remove_region(input_path, output_path, x1, y1, x2, y2, radius=5):
"""Remove rectangular region using inpainting."""
img = cv2.imread(input_path)
h, w = img.shape[:2]
mask = np.zeros((h, w), dtype=np.uint8)
cv2.rectangle(mask, (x1, y1), (x2, y2), 255, -1)
result = cv2.inpaint(img, mask, radius, cv2.INPAINT_TELEA)
cv2.imwrite(output_path, result)
# Example: remove region at coordinates
remove_region('/mnt/user-data/uploads/img.png',
'/mnt/user-data/outputs/clean.png',
x1=700, y1=650, x2=800, y2=720)
```
By Corner
```python
def remove_corner_logo(input_path, output_path, corner='bottom_right',
w_ratio=0.1, h_ratio=0.1, padding=10):
"""Remove logo from corner. corner: top_left, top_right, bottom_left, bottom_right"""
img = cv2.imread(input_path)
h, w = img.shape[:2]
lw, lh = int(w w_ratio), int(h h_ratio)
coords = {
'bottom_right': (w - lw - padding, h - lh - padding, w - padding, h - padding),
'bottom_left': (padding, h - lh - padding, lw + padding, h - padding),
'top_right': (w - lw - padding, padding, w - padding, lh + padding),
'top_left': (padding, padding, lw + padding, lh + padding)
}
x1, y1, x2, y2 = coords[corner]
mask = np.zeros((h, w), dtype=np.uint8)
cv2.rectangle(mask, (x1, y1), (x2, y2), 255, -1)
result = cv2.inpaint(img, mask, 5, cv2.INPAINT_TELEA)
cv2.imwrite(output_path, result)
# Example: remove bottom-right logo
remove_corner_logo('/mnt/user-data/uploads/img.png',
'/mnt/user-data/outputs/no_logo.png',
corner='bottom_right', w_ratio=0.08, h_ratio=0.08)
```
Find Coordinates
```python
img = cv2.imread(input_path)
h, w = img.shape[:2]
print(f"Size: {w}x{h}")
# Gemini λ³ λ‘κ³ λ λ³΄ν΅ μ΄λ―Έμ§ μ°νλ¨ λͺ¨μ리μμ μ½κ° μμͺ½μ μμΉ
# μΌλ°μ μΈ μ’ν: x1=w-150, y1=h-100, x2=w-130, y2=h-55
# μ νν μμΉλ μ΄λ―Έμ§λ§λ€ λ€λ₯΄λ―λ‘ μ‘°μ νμ
```
Output
Always save to /mnt/user-data/outputs/ and use present_files tool.
Notes
- Inpainting works best for small areas with uniform backgrounds
- Gemini logo is typically in bottom-right corner
- Adjust coordinates/ratios based on actual logo position and size
More from this repository10
Generates high-converting landing pages with exceptional design using Next.js 14+, ShadCN UI, and a proven 11-element conversion framework that creates memorable brand experiences.
Automatically documents AI code changes and provides real-time web-based visualization of code modifications.
Generates structured, customizable slash commands for Claude Code, enabling quick and precise AI interactions tailored to specific project needs.
Analyzes project context to automatically expand simple user requests into detailed, comprehensive requirements for more precise AI interactions.
Generates Instagram-style card news series by creating visually appealing 600x600 cards with user-specified topic, colors, and optional background images.
flutter-init skill from bear2u/my-skills
Generates structured, high-quality design prompts for AI image generation by analyzing project context and design requirements.
Generates comprehensive, conversion-optimized landing page design and content strategies with actionable recommendations for creating visually appealing and high-performing web pages.
Orchestrates a dual-AI engineering workflow where Claude Code plans and implements while Codex validates and reviews code iteratively.
workthrough skill from bear2u/my-skills