Individual solutions

We configure image processing systems that are tailored to your individual demands. A selection of available technologies is outlined below.

Tracking

Objects in video images can be recognized through a variety of features: motion, color, size, shape and more. The picture above shows an example application where an object of certain size and color is to be followed.

Camera Calibration

Whether you want to create orthophotos or require a highly accurate, bundle adjustment based camera calibration for your computer vision application, we have the right solution for you.

We offer two different types of calibration approaches:

• Distortion parameters according to Brown: Radial symmetric distortion A1, A2, A3 (A4 on request), radial-asymmetric and tangential distortion (B1, B2) and affinity and shear (C1, C2)
• Advanced calibration routines: we offer an extended calibration approach for high precision applications as a DLL.

Camera calibration can be performed over the Internet. You take pictures according to our specificationa and we perform the calibration via the internet.
We provide a software routine as a DLL to transform your image coordinates into non-distorted image coordinates based on the calibration parameters. Take advantage of the opportunities and optimize the performance of your image processing application.

Subpixel Feature Extraction

To achieve highest accuracy in measurement applications we employ sub-pixel algorithms, e.g. for the determination of ellipse centers. The ellipse center can be determined with an accuracy of 0.03 pixels and better.

Bundle adjustment

Bundle adjustment is an optimization process according to the method of least squares, the "gold" standard for high quality results. We use this method for 2D and 3D tasks. The picture below shows the typical structure of the normal equation matrix with hundreds of unknown values.