Embedded Files
DoEverything3D, Software Tool to Extract Automatically 3D Model of Objects using a Kinect Sensor
3D reconstruction technology is widely used in several applications, some of these are found in media, film, videogames, industries and teaching. Unfortunately, the use of these devices requires some computer knowledge and the acquisition costs are expensive (~ 2000 USD).
DoEveryThing3D is an affordable alternative with an intuitive interface for automated extraction of 3D models of objects by using the Microsoft Kinect sensor. The design of DoEverything3D started mainly with the gather from state of art about analysis and generation of 3D models.
On the other hand, for the validation of DoEverything3D a protocol of integration and quantitative tests was developed. The integration tests were successful, evidenced by the correct implementation of each system requirements. The quantitative tests verified the good quality of the 3D models generated by the system, getting a mean error at 0.3988 mm with a standard deviation (5σ) of 8.846 mm.
Actually, DoEverything3D is a registered software legally recognized by the Interior Ministry of Colombia. This document can be downloaded here.
DoEverything3D uses a rotating platform to scan objects. This rotating platform has the following functionalities:
It communicates with DoEverything3D to respond to any rotating commands.
It rotates full 360° the object to scan.
It communicates to DoEverything3D using USB.
It can rotates the object to scan at specific angle step.
The figure at right shows the configuration module of DoEverythnig3D, it is used to configure the following parameters:
The angle resolution to rotate the object to scan.
The pitch angle of the Kinect sensor.
To Select the communication port with the rotation platform.
To capture a sample range image from the Kinect sensor.
To select the color palette which be applied to the range image.
Once the configuration parameters are introduced, an optional step is to calibrate the Kinect sensor. This is shown in the figure at left. Here users can perform the following functionalities:
User can start the calibration process once the calibration plane is placed at the desired distance.
It is required 10 calibration shots.
Then users can start the calculation of the calibration parameters.
The GUI also shows the current calibration parameters.
Once the new calibration parameters are computed, they are shown in the GUI.
Then users can select if these parameters are saved or not.
Then, once the Kinect sensor is calibrated, DoEverything3D can start the scanning procedure of the interest object. This is done using the GUI shown in the figure at right. Here, users can do the following:
Start the data acquisition.
Users can observe the scanning process graphically.
At the end of the data acquisition, the range image registration starts.
The result of this registration process can be observed by users.
Then users can observe the resulting point cloud, or the surface model.
Finally users can save the results or delete them.
DoEverything3D uses a SQL data base (MySQL) to save all data in each session performed by users. Using this persistence resource, DoEverything3D can issue a report of the user session. This is show in the figure at left, where users can do the following:
Selecting the session to issue a report.
Visualizing the resulting point cloud of the session selected.
Visualizing the resulting surface model of the session selected.
Export all results in a PDF document.
The following video shows how DoEveryThing3D works:
This software was developed for industrial service and academic use. If you like to perform a field test, please contact me:
Prof. Bladimir Bacca Cortes Ph.D.
Address: Cra. 100, Street 13, Universidad del Valle, Melendez, Building 354, Office 2006.
Tel: +5723212100 Ext. 7656
Page updated
Report abuse