Deep Learning
With HALCON 17.12 users are able to train their own classifier using CNNs (Convolutional Neural Networks) based on two pretrained networks that are included in HALCON. These have been highly optimized for industrial applications and are based on hundreds of thousands of images.
During training, HALCON automatically learns which features can be used to identify different classes – a big advantage compared to all previous classification methods. This significantly reduces programming efforts.
After training the CNN, it can be used for classifying new data with HALCON.
Deflectometry
HALCON 17.12 includes a new method for inspecting specular reflecting surfaces to detect defects like dents or scratches which can hardly be identified using conventional surface inspection techniques. Deflectometry uses specular reflections by observing mirror images of known patterns and their deformations on the surface.
Easy Code Export
HDevelop provides a new library export that makes the use of HALCON procedures from C++ as easy and intuitive as calling any other C++ function. This is possible via a C++ wrapper that encapsulates the necessary HDevEngine API calls. This new library export also generates CMake projects which can easily be configured to output project files for many popular IDEs, such as Visual Studio.
The new functionality is accessible from the HDevelop GUI and via command line interface.
Improved Automatic Text Reader
HALCON 17.12 features an improved automatic text reader, which now detects and separates touching characters more robustly.
3D Fusion
HALCON now offers a new method that fuses multiple 3D point clouds into one watertight surface. This new method is able to combine data from different or even various 3D sensors, like a stereo camera, time of flight camera, or fringe projection. The data from these sensors get fused into one highly optimized 3D point cloud. This technology is especially useful for reverse engineering.
GigEVision2
HALCON now provides the new GigEVision2 interface which supports devices complying with the GigE Vision 2.x standard and replaces the old GigEVision interface. In particular, this new interface supports the transmission of chunk data and additional payload types including the multipart payload. In combination with a 3D sensor that supports the multipart payload type the GigEVision2 interface is now able to directly create an ObjectModel3D when using grab_data or grab_data_async.