FOVEX 3D/Creating Project Manual

From FOVEX Help
Jump to: navigation, search

Setting up a Project

The following steps show how to setup a measurement project manually:

Step 1. Create a new project, File > New >New Project. The project name by default is “project0” In Project Explorer, right click on the “project0” and Reload Project.


Create a new solution. Add a new project to the solution.
Create a new solution. Add a new project to the solution.
Step 2. Add a camera, right-click on the frame Array Camera in the Project Explorer and click on Add Camera. Sensor is added to the Project Explorer.

If your project consists of more than one camera, you can add them by repeating Add Camera.

Addition of a Camera to the project. Project Explorer after addition of the camera.
Addition of a Camera to the project. Project Explorer after addition of the camera.
Step 3. Add images to your sensor, right-click on the Images under your Camera and click on Add Image(s) > Add Image(s) Browse and select the image files that belong to this camera.

A dialog box appears to fill in the camera setting parameters. Press OK. By clicking on Images you will have the list of all project images in Report View.


Add images to the project.
Add images to the project.
Automated sensor parameter filling. The approximate value of Camera constant has been given here.
Automated sensor parameter filling. You need to provide the correct value of the parameters, if the value of the parameters under the New column is 0. Here in this figure, the value of Camera constant is 0 and we should provide an approximate value for it. The approximate value of Camera constant has been given here. Pressing OK will add the images in the project.
The Main Window of software after project setup.
The Main Window of software after project setup.
Note: By clicking on Setting of the Sensor, the camera parameter setting at Report View is displayed. The values for the parameters of the yellow field have been already filled in by Auto Filling Sensor Parameters in the previous Step 3.

For an accurate 3D measurement all camera parameters indicated should be estimated preferly with a separate step only for calibrating the camera. Page “Frame Array Camera Calibration” describes how to calibrate the camera accurately.

Page “Frame Array Camera Calibration” describes how to calibrate the camera accurately.
Step 4. Up to this step the essential part of the project has been configured. The project can be saved on hard disk. Right click on the name of project which is currently “project0”:
Up to this step the essential part of the project has been configured.


Step 5. Up to this step the essential part of the project has been configured. However, if additional data are available, they can be imported into the project. The following additional data can be imported through Import menus which are accessed by a right-click on relevant items at Project Explorer:
Reference points.
Control points and check points.
Reference points. Control points and check points.
Camera parameters. Image points.
Camera parameters. Image points.
Exterior Orientation Parameters of the sensor. List of scale-bars.
Exterior Orientation Parameters of the sensor. List of scale-bars.

Accessing Project Data and Results

FOVEX 3D Graphical User Interface (GUI) provides 4 different views (tabs) for accessing data and results:

FOVEX 3D Graphical User Interface (GUI) provides 4 different views (tabs) for accessing data and results


Image View It displays the image content and related measurement data. To view an image, double-click on the Image item in the Project Explorer.
Network View It displays the 3D measurement network including camera stations, 3D points, error vectors and CAD objects.
Report View It displays the content of the item at Project Explorer in tabular and text forms. The tabular or text forms of the items at Project Explorer are shown in Report
Layer View It displays the defined layers for the project. To view content, click on Layer View tab.

Image Orientation

Image orientation means finding the most accurate estimate for three orientation angles (w, phi, and kappa) and the 3D coordinates of the location of the camera stations. These 6 parameters are called exterior orientation parameters of an image.

Estimation of these parameters does not require external information such as distance between camera stations or the location of the image points in 3D.

There are two automated ways and one manual way to perform Build Network with FOVEX 3D. The following sections describe two typical ways of performing Build Network.

Automated with Coded Targets

If PhotoCore coded targets are available in your images, this method for image orientation is suggested.

An image with PhotoCore coded targets. PhotoCore coded target.
An image with PhotoCore coded targets. PhotoCore coded target.

This method is fully automated and is the most reliable and accurate way for image orientation. The following steps provide necessary information to orient images successfully by using coded targets.

Step 1. From the toolbar press on Auto Process with Coded Targets (Auto Process toolbar).

Auto Process with Coded Targets


This process includes camera self-calibration for the selected free additional parameters.

Step 2. Scale of the network can be recovered by using the size of coded targets:
Choose “Add Coded Target Scale Bars” Scale of the network can be recovered by using the size of coded targets

Choose “Add Coded Target Scale Bars”.


The result of added scale-bars to the network.

The result of added scale-bars to the network.


Network View after image orientation with coded targets and scaling with coded target scale bar.

Network View after image orientation with coded targets and scaling with coded target scale bar.


Note: The group of Automated Target Detection parameters influences the results of coded target recognition. This group of parameters can be accessed from Project Explorer the Settings of the project:
Automated Target Detection


Automated with Natural Features

If PhotoCore coded targets are not available, the method based on natural features for image orientation is suggested which is fully automated. The quality of image orientation depends on the quality of the extracted features in the image. The following steps provide necessary information to orient images successfully.

Step 1. From the toolbar press on Auto Process with Natural Features (Auto Process toolbar).
Auto Process with Natural Features
This process includes camera self-calibration for the selected free additional parameters. It should be noted that the quality of camera calibration depends on the goodness and number of natural features.


The result of Auto Process with natural features
The result of Auto Process with natural features. The network does not have yet a proper scale.
Step 2. Scale of the network can be recovered by manual scale-bar definition.

This process has been explained in the section “Add Scale Bar”.


Note: The group of Feature Detection & Matching parameters influences the results of feature detection and matching. This group of parameters can be accessed from Project Explorer the Settings of the project:
Feature Detection & Matching


Manual with Image Point Measurement

Note: If performing the Image Orientation with manual image point measurement, the camera additional parameters should be fixed in advance to avoid performing on-the-fly self-calibration. The following shows that all parameters are fixed to the Given values:
The following shows that all parameters are fixed to the Given values


This method needs the interaction and supervision of the user. It is suggested for the projects in which PhotoCore Coded Targets or good quality of natural features are not available. The following steps provide necessary information to orient images by manually measurement of the image points:

Step 1. Open two images which have sufficient (>50%) overlap from the list of images at Project Explorer. These two images should not be taken from an identical location.
Step 2. Measure at least five corresponding image features among these opened images and associate unique but corresponding labels to them. Choose your image points such to spread them across the image. Image features can be measured by using the “Draw manual point” at the toolbar of the Image View:

Draw manual point


Correspondence of image feature measurement is defined in FOVEX 3D by giving a unique label to the corresponding features. For example, “A0” in the following:

Correspondence of image feature measurement is defined in FOVEX 3D by giving a unique label to the corresponding features. For example, “A0” in the following


The following Figure shows all the manually measured corresponding image points at two different images:

The following Figure shows all the manually measured corresponding image points at two different images

Step 3.

Perform Orient Images. Orient Images can be accessed from the Adjustment toolbar:

Orient Images

After a successful image orientation, the decoration symbol of the image at Project Explorer will become green:

After a successful image orientation, the decoration symbol of the image at Project Explorer will become green.


If the images are not oriented, you need to measure more corresponding image features similarly as explained at step 2.

Step 4. If the number of corresponding image points is more than 12 and still the images are not

oriented, go to step 5. Before going to step 5, double check if all image points are corresponding to each other.

Step 5. If all images are oriented or you would like to stop building network process, then go to 7.
Step 6. Open the next non-oriented image (with a red decoration color) and go to Step 2.
Step 7. End of image orientation. The following figure shows the result of image orientation for this project which was consisted of 4 images:


The following figure shows the result of image orientation for this project which was consisted of 4 images

Automated Labeling

Image points which do not have a label (for example measured circular targets) can be labeled automatically by clicking on Automated Labeling from the tool bar Adjustment:

Automated Labeling


Note: Automated labeling is usually used for industrial project when the object surface is featured with retro-reflective or paper circular targets.


The Following figures show the result of automated target labeling:

The Following figures show the result of automated target labeling

The circular black targets on the box are labeled automatically. The pre-fix “Auto” is added to the label of Auto-labeling process.


The Network View (3D View) of the automated target labeling.

Scaling the Project

The final step of the creation of a 3D measurement project is scaling the project. Scale-bars can be added to the project in three ways. Click on Scale Bars at Project Explorer and go to the tab of Report View. Right Click on the middle of window and select Add Scale Bar. The next sections describe possible ways to adding scale to project.

Context menu of Scale Bars at Report View: the addition of a scale bar to the project.


Add Scale Bar

By clicking on Add Scale Bar a tabular form is displayed as follows:

By clicking on Add Scale Bar a tabular form is displayed as follows


The following information should be given for each scale bar:

  • A unique ID of the scale-bar
  • The first ID of scale-bar end point 1
  • The second ID of scale-bar end point 2
  • Calibrated temperature, if available
  • Calibrated length of the scale-bar
  • Standard deviation (uncertainty) of the scale-bar length


Click on the Active Check Box and the project will be scaled to the size of the given scale bar.


Add Coded Target Scale Bar

By clicking on Add Coded Target Scale Bars the following dialog box is displayed:

By clicking on Add Coded Target Scale Bars the following dialog box is displayed


Provide the size of coded target which is written at the corner of the coded target and Press Add. The project will be scaled automatically with the given length of scale-bars.

Size of Coded Target is the above indicated distance between the two circular targets.