nvJPEG-Decoder-Backend-ROI

nvJPEG-Decoder-DifferentBackEnd

Description

The sample codes shows the usage of decouple nvJPEG API on decoding JPEG images. Extra functionalities including different backends and ROI are supported.

Key Concepts

Image Decoding, different backends from NVJPEG Library

Supported SM Architecture

SM 3.0 SM 3.5 SM 3.7 SM 5.0 SM 5.2 SM 6.0 SM 6.1 SM 7.0 SM 7.2 SM 7.5 8.0

Supported OS

Linux, Windows

Supported CPU Architecture

x86_64

CUDA APIs involved

NVJPEG

Prerequisites

• A Linux/Windows system with recent NVIDIA drivers.
• Install the CUDA 11.2 Update 1 toolkit

Building

Build command

$ mkdir build
$ cd build
$ cmake ..
$ make

Usage

Usage: ./nvJPEGROIDecode -i images_dir [-roi roi_regions] [-backend backend_enum] [-b batch_size] [-t total_images] [-w warmup_iterations] [-o output_dir] [-pipelined] [-batched] [-fmt output_format]
Parameters: 
        images_dir      :       Path to single image or directory of images
        roi_regions     :       Specify the ROI in the following format [x_offset, y_offset, roi_width, roi_height]
        backend_eum     :       Type of backend for the nvJPEG (0 - NVJPEG_BACKEND_DEFAULT, 1 - NVJPEG_BACKEND_HYBRID,
                                2 - NVJPEG_BACKEND_GPU_HYBRID)
        batch_size      :       Decode images from input by batches of specified size
        total_images    :       Decode this much images, if there are less images 
                                        in the input than total images, decoder will loop over the input
        warmup_iterations       :       Run this amount of batches first without measuring performance
        output_dir      :       Write decoded images as BMPs to this directory
        pipelined       :       Use decoding in phases
        batched         :       Use batched interface
        output_format   :       nvJPEG output format for decoding. One of [rgb, rgbi, bgr, bgri, yuv, y, unchanged]

Example 1 - Choosing different backend

Default Backend
Command: ./nvJPEGROIDecode -i ../input_images/ -backend 0

...
...
Total decoding time: 0.0136991 (s)
Avg decoding time per image: 0.00114159 (s)
Avg images per sec: 875.972
Avg decoding time per batch: 0.00114159 (s)

Hybrid BackEnd
Command: ./nvJPEGROIDecode -i ../input_images/ -backend 1

...
...
Total decoding time: 0.0138813 (s)
Avg decoding time per image: 0.00115678 (s)
Avg images per sec: 864.47
Avg decoding time per batch: 0.00115678 (s)

GPU Backend
Command: ./nvJPEGROIDecode -i ../input_images/ -backend 2

...
...
Total decoding time: 0.0166718 (s)
Avg decoding time per image: 0.00138931 (s)
Avg images per sec: 719.779
Avg decoding time per batch: 0.00138931 (s)

Example 2 - Decode with ROI

Command: ./nvJPEGROIDecode -i ../input_images/ -roi 0,0,64,64

Original image:


Result: