Logo
Sign in
Product Logo
NVIDIA nvImageCodecNVIDIA

The nvImageCodec is a library of accelerated codecs with a unified interface. It is designed as a framework for extension modules that deliver codec plugins. The library supports GPU-accelerated image processing codecs, including nvJPEG, nvJPEG2000, and nvTIFF, along with fallback options to provide comprehensive support for CPU-based codecs.

Vendor

Vendor

NVIDIA

Company Website

Company Website

Product details

NVIDIA nvImageCodec is a library of accelerated codecs designed to provide a unified interface for GPU-accelerated image processing. It supports a variety of image formats and offers high-performance decoding and encoding capabilities. The library is built to handle large, complex image datasets efficiently, leveraging NVIDIA's CUDA platform for optimal performance.

Features

  • Unified API: Provides a unified API for decoding and encoding images, simplifying integration and usage.
  • Batch Processing: Supports batch processing with variable shape and heterogeneous formats, enhancing efficiency.
  • Codec Prioritization: Automatically prioritizes codecs with fallback options to ensure comprehensive support.
  • Built-in Parsers: Includes built-in parsers for image format detection, supporting JPEG, JPEG 2000, TIFF, BMP, PNG, PNM, WebP.
  • Python Bindings: Offers Python bindings for seamless integration with CV-CUDA, PyTorch, and CuPy.
  • Zero-Copy Interfaces: Provides zero-copy interfaces to CV-CUDA, PyTorch, and CuPy, optimizing memory usage.
  • End-to-End Acceleration: Includes end-to-end accelerated sample applications for common image transcoding tasks.

Benefits

  • Efficiency: Enhances processing efficiency by supporting batch processing and zero-copy interfaces.
  • Flexibility: Offers extensive support for various image formats and seamless integration with popular frameworks.
  • Performance: Leverages GPU acceleration to provide high-performance decoding and encoding capabilities.
  • Scalability: Handles large, complex image datasets efficiently, making it suitable for demanding applications.