Module ev_filtering 

Event filtering module for evlib

This module provides comprehensive event filtering functionality for event camera data. It supports both streaming and batch processing with a focus on performance and memory efficiency.

Features

• Temporal filtering: Time-based event selection with microsecond precision
• Spatial filtering: Region of interest (ROI) and coordinate-based filtering
• Polarity filtering: Positive/negative event separation
• Hot pixel removal: Statistical outlier detection and removal
• Noise filtering: Refractory period and distance-based denoising
• Event downsampling: Intelligent event reduction strategies
• Pixel dropping: Selective pixel masking and exclusion

Performance

All filtering operations are optimized for:

• Vectorized processing using SIMD instructions
• Memory-efficient streaming for large datasets
• Parallel processing using Rayon
• Zero-copy operations where possible

Usage

use evlib::ev_filtering::{FilterConfig, filter_events, TemporalFilter, SpatialFilter};
use evlib::{Event, Events};

// Create sample events
let events = vec![
    Event { t: 1.0, x: 100, y: 200, polarity: true },
    Event { t: 2.0, x: 150, y: 250, polarity: false },
    Event { t: 3.0, x: 200, y: 300, polarity: true },
];

// Configure filters
let config = FilterConfig::new()
    .with_temporal_filter(TemporalFilter::new(1.5, 2.5))
    .with_spatial_filter(SpatialFilter::roi(120, 180, 220, 280));

// Apply filters
let filtered_events = filter_events(&events, &config)?;

Re-exports

pub use config::FilterConfig;

pub use config::FilterError;

pub use config::FilterResult;

pub use denoise::DenoiseFilter;

pub use denoise::RefractoryFilter;

pub use downsampling::DownsamplingFilter;

pub use downsampling::DownsamplingStrategy;

pub use drop_pixel::DropPixelFilter;

pub use drop_pixel::PixelMask;

pub use hot_pixel::HotPixelDetector;

pub use hot_pixel::HotPixelFilter;

pub use polarity::PolarityFilter;

pub use spatial::CircularROI;

pub use spatial::MultipleROIs;

pub use spatial::Point;

pub use spatial::PolygonROI;

pub use spatial::ROICombination;

pub use spatial::RegionOfInterest;

pub use spatial::SpatialFilter;

pub use temporal::TemporalFilter;

Modules

config

Filter configuration and error handling

denoise

Polars-first noise filtering operations for event camera data

downsampling

Polars-first event downsampling operations for reducing event density

drop_pixel

Polars-first pixel dropping and masking operations for event camera data

hot_pixel

Polars-first hot pixel detection and removal for event camera data

polarity

Polars-first polarity filtering operations for event camera data

presets

Convenience function for common filtering operations

python

Python bindings for event filtering functionality

spatial

Polars-first spatial filtering operations for event camera data

temporal

Polars-first temporal filtering operations for event camera data

utils

Utility functions and statistics for event filtering using Polars DataFrames

Structs

FilterStats

Statistics about filtering operations

Constants

COL_POLARITY

COL_T

COL_X

COL_Y

Functions

filter_events_dataframe

Polars-first filtering function that applies filters entirely using LazyFrame operations