Optimize Performance

This guide covers techniques to improve Quagga2's barcode scanning performance.

Overview

Performance optimization in Quagga2 involves balancing accuracy against speed. The key areas to optimize are:

Input resolution and scaling
Locator configuration
Reader selection
Processing frequency

Input Resolution

Using `inputStream.size`

Reducing the processing resolution is the most effective way to improve performance:

Quagga.init({
  inputStream: {
    size: 640  // Process at 640px max dimension instead of full resolution
  }
});

Recommended values:

1280px - High quality, slower (good for static images)
800px - Balanced (default for decodeSingle)
640px - Fast (recommended for live scanning)
480px - Very fast (may reduce accuracy)

Camera Constraints

Request only the resolution you need:

inputStream: {
  constraints: {
    width: { ideal: 1280 },
    height: { ideal: 720 }
  }
}

Locator Configuration

Half Sampling

Keep halfSample: true (default) for faster localization:

locator: {
  halfSample: true  // Processes at half resolution
}

Patch Size

Larger patch sizes are faster but may miss small barcodes:

locator: {
  patchSize: "large"  // Options: x-small, small, medium, large, x-large
}

Reader Selection

Only enable the barcode formats you need:

decoder: {
  readers: ["code_128_reader"]  // Don't enable all readers
}

Processing Frequency

Limit scan rate to reduce CPU usage:

Quagga.init({
  frequency: 10  // Max 10 scans per second
});

Disable Localization

If barcode position is fixed, disable localization entirely:

Quagga.init({
  locate: false,
  inputStream: {
    area: {
      top: "25%",
      right: "25%",
      bottom: "25%",
      left: "25%"
    }
  }
});

Configuration Reference - All configuration options
How Barcode Localization Works - Understanding the algorithm

← Back to How-To Guides

2.4 KiB Raw Blame History