Computer Vision Technology for Star Trek Online Builds
SISTER is a powerful computer vision system that automatically recognizes equipment and traits in Star Trek Online screenshots. Using classical computer vision techniques combined with perceptual hashing and SSIM matching, SISTER can extract detailed build information from your screenshots, making it easier to analyze and share your builds with the community.
🚀 Key Features
OCR-based Label Detection
Currently uses EasyOCR with optional GPU acceleration for fast and accurate text recognition in screenshots.
Icon Slot Extraction
Advanced contour analysis and smart filters precisely locate equipment and trait icons in your screenshots.
Intelligent Icon Matching
Combines perceptual hashing with SSIM matching for highly accurate equipment and trait identification.
Quality Detection
Automatically identifies quality overlays (Epic, Very Rare, etc.) on equipment icons.
Multi-Platform Support
Works with screenshots from PC and Console versions, handling both space and ground builds.
Extensible Design
Modular architecture allows for easy customization and integration into other tools.
Get Started with SISTER
Open Source Project
SISTER is completely open-source and available on GitHub. Join our community to report issues, suggest features, or contribute to development.