skills.rest
limit5
limit5COMMUNITY
Explore all skills by this author →
GitHub

npu-barcode

Detect and decode barcodes and QR codes on NPU-enabled embedded cameras.

1|Updated Apr 9, 2026
Search Tags:#embedded systems#decoding#qrcode#barcode#npu#zbar#hybrid detection
One-click install
npx skills add https://github.com/limit5/OmniSight-Productizer --skill npu-barcode
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: npu-barcode
Source: https://github.com/limit5/OmniSight-Productizer/tree/main/configs/skills/npu-barcode
Command: npx skills add https://github.com/limit5/OmniSight-Productizer --skill npu-barcode

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Traditional barcode and QR code scanning on embedded AI camera systems suffers from high latency and low reliability, especially with angled, blurry, or multiple codes per frame. This Skill eliminates that bottleneck by combining NPU-accelerated region localization with proven traditional decoder fallback for consistent, fast scan performance.

Core Features & Use Cases

  • NPU-Accelerated Localization: Uses lightweight YOLO-tiny or SSD-MobileNet models on NPU to quickly identify barcode regions in camera frames, minimizing processing time.
  • Hybrid Decoder Fallback: Crops detected regions and feeds them to ZBar (for 1D/QR codes) or ZXing (for DataMatrix) for reliable decoding, with full-frame fallback if no regions are detected by the NPU.
  • Use Case: Ideal for embedded AI camera deployments in warehouses, retail, or access control systems that need to scan asset tags, product QR codes, or entry passes in real time, even when codes are partially obscured or viewed at an angle.

Quick Start

Use the npu-barcode skill to detect and decode all barcodes and QR codes from the live RTSP feed of the warehouse security camera system.

Frequently Asked Questions about npu-barcode

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I decode barcodes and QR codes on an embedded AI camera with low latency?▼

To achieve low-latency barcode decoding on embedded systems, offload region localization to NPU hardware using lightweight models, then crop regions for traditional decoders. This hybrid approach minimizes processing time for real-time edge camera deployments.

Does NPU-accelerated scanning work reliably on angled or low-quality QR codes?▼

NPU-accelerated scanning handles angled and low-quality QR codes by using hardware localization to identify regions, then applying ZBar or ZXing fallback decoders. If no regions are detected, a full-frame decoding fallback ensures consistent reliability.

What's the best way to track warehouse inventory using RTSP camera feeds?▼

Tracking warehouse inventory via RTSP feeds requires real-time scanning of asset tags and product QR codes. Utilizing NPU hardware for region localization combined with traditional decoder fallback delivers high decode success rates for demanding edge computing applications.

Can I use ZBar and ZXing together for DataMatrix and 1D barcode detection?▼

Yes, hybrid detection uses ZBar for 1D and QR codes alongside ZXing for DataMatrix formats. This combination ensures reliable decoding across multiple barcode types by leveraging traditional decoders after NPU-accelerated region localization.

Why does traditional barcode scanning suffer from high latency on edge devices?▼

Traditional barcode scanning on edge devices suffers from high latency due to full-frame processing overhead. Offloading region localization to NPU hardware with lightweight YOLO-tiny or SSD-MobileNet models eliminates this bottleneck for real-time performance.

Related Skills

Explore More Software Engineering Skills→
limit5limit5

SKILL_HD_SENSOR_SWAP_FEASIBILITY

Assess alternative image sensor compatibility with embedded AI camera systems.

Community
Basic
adiytharpansaadiytharpansa

barcode-scanner

Scan barcode images to retrieve product details and pricing information.

Community
Intermediate
SharpAISharpAI
3.0k

yolo-detection-2026

Detect objects in video frames using YOLO 2026 models.

Official
Advanced
ovachieverovachiever
52

llama-cpp

Run LLM inference on CPU and non-NVIDIA hardware with GGUF quantization.

Community
Advanced