nvJPEG2000 is a high-performance GPU-accelerated library designed for decoding JPEG 2000 format images. It is tailored for application developers and researchers working with JPEG 2000 formatted images in fields such as deep learning, medical imaging, digital pathology, remote sensing, and digital cinema applications. The library reads and decodes JPEG 2000 image data from CPU memory, with the decoded output in GPU memory, leveraging both CPU and GPU for efficient decoding.

Features

• Output Formats: Supports grayscale and color images with arbitrary width and height.
• Compression Techniques: Provides both lossy (wavelet CDF 9/7) and lossless (wavelet CDF 5/3) image compression and decompression.
• File Format Support: Compatible with jp2 file format and JPEG 2000 code stream.
• High Performance: Achieves up to 4x faster lossless decoding and 7x faster lossy decoding compared to CPU-only methods.

Benefits

• Enhanced Performance: Delivers significant performance improvements for decoding JPEG 2000 images, making it suitable for high-throughput applications.
• Versatility: Applicable to various fields, including deep learning, medical imaging, digital pathology, remote sensing, and digital cinema.
• Efficiency: Reduces processing time and improves overall efficiency by leveraging GPU acceleration.