VCA

An open source video complexity analyzer for predicting spatial and temporal complexity at frame, segment, and video level.

Features:

• Spatial complexity (E) and temporal complexity (h) per frame
• Brightness (L) measurements
• Chroma statistics for U and V components
• Shot detection with CSV output
• Per-block results exportable to YUView visualization format
• Configurable block sizes (8/16/32 pixels)
• SIMD-optimized for performance

Use cases:

• Per-title encoding parameter optimization
• Adaptive video streaming workflows
• Green video complexity analysis for efficient encoding

Input formats:

• Raw YUV
• Y4M
• Pipe from FFmpeg via stdin

Installation:

Build from source using CMake, or use the provided Docker container.

Documentation: https://cd-athena.github.io/VCA/