Image Decode Implementation with NVJpeg and CV2

author: Junjie Jiang

Description

An image decode operator implementation with OpenCV and NvJpeg. The decoder will use GPU to decode jpeg images.

Only supports linux.

Code Example

Load a image from path './dog.jpg'.

Write the pipeline in simplified style:

import towhee

towhee.glob('./dog.jpg') \
  .image_decode.nvjpeg() \
  .show()

Write a same pipeline with explicit inputs/outputs name specifications:

import towhee

towhee.glob['path']('./dog.jpg') \
  .image_decode.nvjpeg['path', 'img']() \
  .select['img']() \
  .show()

Factory Constructor

Create the operator via the following factory method:

image_decode.nvjpeg()

Interface

An image decode operator takes an image path as input. It decodes the image back to ndarray.

Parameters:

img: str

​ Image file path.

Returns: towhee.types.Image (a sub-class of numpy.ndarray)

​ The decoded image data as numpy.ndarray.