Skip to content

v0.2.0 - Major Refactor

Pre-release
Pre-release

Choose a tag to compare

View all tags
@qhuboo qhuboo released this 21 Mar 04:58
· 4 commits to main since this release
v0.2.0
e240dfe
This commit was created on GitHub.com and signed with GitHub’s verified signature.
GPG key ID: B5690EEEBB952194
Verified
Learn about vigilant mode.

BREAKING CHANGES - Major architectural refactor from v0.1.24

Overview

This release represents a complete overhaul of WunderScout's core architecture. Over 50% of the codebase has been refactored to improve maintainability, performance, and API clarity.

This is an experimental release and should NOT be used in production environments.

What's New

  • Restructured core detection pipeline
  • Improved heatmap generation with histogram and KDE support
  • Enhanced team classification system
  • Better coordinate mapping with persistent homography
  • Comprehensive CSV export functionality
  • New public API with cleaner interfaces

Breaking Changes

  • Complete API restructure - not compatible with v0.1.24
  • New Frames class for detection results
  • Updated HeatmapGenerator interface
  • Changed configuration patterns
  • Modified CSV export format

Public API

The library now exports:

  • Models - Model management
  • Detector - Video processing
  • Frames - Detection results
  • HeatmapGenerator - Heatmap generation
  • Heatmap - Heatmap data container
  • SaveResult - Save operation results
  • ClassId - Detection class enumeration

Known Issues

  • Core detection works but is extremely slow and needs major optimization work
  • Annotation rendering needs improvements
  • KDE heatmaps require minimum sample sizes
  • Team classification may struggle with similar colors

Installation

pip install wunderscout==0.2.0

Documentation

See README.md for complete API documentation and usage examples.

Note: This is a pre-release version. Please report issues on GitHub.

Assets 2
Loading