Skip navigation links
Open Inventor 9.9

Package com.openinventor.imageviz

The ImageViz extension allows easy integration of advanced 2D and 3D image processing and analysis capabilities into imaging software applications.

See: Description

Package com.openinventor.imageviz Description

The ImageViz extension allows easy integration of advanced 2D and 3D image processing and analysis capabilities into imaging software applications. It provides an extensive collection of high-performance parallelized 2D/3D image processing and analysis operators to implement complete application workflows, including image filtering and preprocessing, segmentation, and data analysis. ImageViz allows developers to build automated workflows thanks to ImageViz engine objects.

Introduction to ImageViz

Specifically designed for application developers, the Open Inventor ImageViz software development toolkit (SDK) allows easy integration of advanced 2D and 3D image processing and analysis capabilities into imaging software applications in medical and life sciences, industrial inspection, materials science, and geosciences.

Open Inventor ImageViz SDK provides an extensive collection of high-performance parallelized 2D/3D image processing and analysis operators to implement application workflows, including:

Workflow example

ImageViz is tightly integrated into the Open Inventor SDK and its standard concept of engines, nodes, and fields. Open Inventor data objects can be connected as input to ImageViz processing engines. Moreover, any ImageViz engines output can be connected to Open Inventor or VolumeViz, allowing easy integration into existing applications.

ImageViz integration with Open Inventor

Getting started with ImageViz

This section introduces how to start working with ImageViz from scratch. Each demo below provides an example for performing a step of a classical image processing chain.

Please note when using the ImageViz documentation: At the top of each engine's page there is a link to a page containing general information about the algorithms used in that group of engines.

1 Image Filtering

The pre-processing step helps to enhance the quality of the image thanks to an image processing filter.

First it is shown how to load a 2D image from a file into an SoImageDataAdapter object.

Denoising a 2D image.

01.2 Denoising a 3D image

This example shows how to denoise an image using a bilateral filter.

First it is shown how to load a 3D image from a file into an SoImageDataAdapter object.

Denoising a 3D image.

2 Image Segmentation

Segmentation is a very important step of an image processing workflow. This is where objects to identify in the scene are isolated in a binary or label image.

Once the input image is loaded, the SoLabelingProcessing engine is applied to assign a different label to each object of the scene.

Label analysis.

3.2 Define a custom measure

ImageViz proposes a substancial set of native measures . However it may be necessary to define its own formula to be computed in a label analysis.

This example shows how to define a custom measure.

The workflow is almost the same as the Label analysis example.

Define a custom measure.

3.3 Customize measure attributes extraction

Some of the ImageViz native measures have attributes (measures related to Ferets diameters, histogram computation or cooccurrence matrix).

This example shows how to modify the attributes of a measure.

The Feret's diameter ratio computes the ratio between 2 Feret's diameter. Depending on the angular step used to compute the distribution of Feret's diameters, the precision of the result can change.

The workflow is almost the same as the Label analysis example.

Customize measure attributes.

4 Image Management

This section contains examples for handling input and output data.

4.1 Creating a 2D image

This example demonstrates how to create a new 2D image containing data from a buffer.

First it is shown how to fill an SoCpuBufferObject with synthetic values that would represent a bright square on a gray background.

Then this buffer is set in an SoMemoryDataAdapter that corresponds to a 2D 512x512 image.

At least a simple scene graph is built and the image is displayed as an orthoslice in a viewer.

4.2 Creating a 3D image

This example demonstrates how to create a new 3D image containing data from a buffer.

First it is shown how to fill an SoCpuBufferObject with synthetic values that would represent a bright 3D bar on a gray background.

Then this buffer is set in an SoMemoryDataAdapter that corresponds to a 3D 256x256x32 image.

At least a simple scene graph is built and the image is displayed in a viewer with as a volume rendering.

4.3 Extracting a rectangular ROI

It is sometimes necessary to select a Region Of Interest (ROI) inside an image. This allows removing non relevant data or saving computation time during the validation step of a workflow.

This example shows how to load an image from a file into an SoImageDataAdapter object.

Extracting a rectangular ROI.

4.4 Launch a batch processing

Once an image processing chain has been validated on a set of representative samples it can be useful to apply the same workflow on a stack of images located in a same folder.

This example shows how to apply a gradient operator on a list of images and save the result in a new file.

A list of every TIFF files located in a given directory is defined (/ImageViz/Data).

Launch a batch processing.

Toward a complete workflow

These examples illustrates how a standard image processing workflow can be implemented. Each step is applied from loading a grayscale image to displaying results of an analysis.

All parameters of the workflow engines are exposed in the user interface, the user can change the parameters and verify the result on the output image.

Skip navigation links
Open Inventor 9.9

Generated on September 3, 2019, Copyright Thermo Fisher Scientific. All rights reserved. http://www.openinventor.com