New operators based on new technologies now make machine vision more accurate, more reliable, and faster than ever. With the new version 7.1 of their flagship HALCON, MVTec Software GmbH, Munich (Germany), again demonstrates its unique competence.
The quality of machine vision fluctuates with speed, accuracy, and
reliability. Whereas previously the illumination and measurement setup
were crucial parameters of each application, today software
increasingly becomes the key component for machine vision. First class
software today is able to compensate weaknesses of the hardware or the
setup, which was not possible some years ago. To solve these
challenging tasks, HALCON 7.1 provides more than 1150 operators.
In the tradition of all previous versions, HALCON 7.1 can also be seen
as a protection of investment: HALCON 7.1 is hardware-independent, and
thus, compatible with many operating systems. Furthermore, HALCON 7.1
provides interfaces to more than 50 frame grabbers and hundreds of
industrial cameras.
Innovations in HALCON 7.1
Like previous versions, version 7.1 offers new operators and
technologies that are either new on the world market or at least unique
in terms of speed and accuracy. The most important innovations are
described below:
Data code reader PDF417
PDF417 codes, typically used for Electronic Data Interchange (EDI), are
increasingly popular. The high data capacity of facilitates replacing
paper documents such as manifests or bills of lading. EDI, the linking
of databases between a supplier and customer, provides real-time
information about a shipment to the recipient. A manufacturer can then
know exactly when critical components will be arriving for the assembly
of a product.
- The benefit of HALCON 7.1: HALCON’s PDF417 reader is robust to
occlusions and to poor print quality, invariant to image contrast and
to rotation, and allows simultaneous detection of multiple PDF417
instances.
Automatic Parameter Determination
Matching is a widely used method in machine vision. The advanced
matching (shape- and component-based matching) of HALCON 7.0 was
already unique, but automatic parameter determination now eases the
matching and can be used for following applications: alignment,
completeness checks, inspection, measuring and comparison, position
recognition, print inspection, recognition of objects, and shape
recognition.
- The benefit of HALCON 7.1: The automatic parameter determination can
help new users by needing less time to learn how to use the matching.
Inexperienced programmers can develop their application faster. Line Scan Camera Calibration
HALCON’s 3D camera calibration for area scan cameras for area scan
cameras was already a unique feature. MVTec now offers the same
technology also for line scan cameras under HALCON 7.1. Thus, this
advantageous and frequently used feature is also available where line
scan cameras are used for inspection e.g., in the automotive, print,
semiconductor, food and beverage, wood and timber industry, and other
production processes in which conveyor belts are being used.
- The benefit of HALCON 7.1: With the 3D camera calibration – now also
for line scan cameras, the interior camera parameter and pose are
determined and pixel coordinates can be converted into world
coordinates. Thus, robot control becomes easier. For highly accurate
and flexible measurements, calibration is essential.
Gray Value Calibration
Gray value calibration is a preprocessing step that can lead to higher
accuracy in all subpixel processing methods (point finders, measurements, line and edge extraction).
- The benefit of HALCON 7.1: No complex calibration plate is necessary,
no 3D camera calibration and no optimized illumination is required.
Speedup of Filters and Arithmetic Operations
For MVTec, it is very important to further increase the speed of
machine vision software processing. MMX (Multi Media Extension) is an
instruction set designed by Intel to accelerate the processing of large
integer data sets as of video, audio, or image data files.
- The benefit of HALCON 7.1: Now, all filters using MMX additionally
advance in speed by HALCON 7.1. Now, frequently used filters are faster
by 300% to 400%(!).
Self Calibration
Normally, 3D calibration needs a set of images showing a calibration
plate in various positions and orientations. Self calibration makes
this procedure dispensable. The camera parameters can be calculated
simply by rotating the camera.
- The benefit of HALCON 7.1: Self calibration allows to automatically
derive the camera parameters by just providing several images from an
arbitrary scene in which the camera is rotated around its optical
center. Self calibration also improves mosaicking (image stitching),
because now it is possible even to generate a 360° panorama image.
Uncalibrated Stereo
As calibration data, uncalibrated stereo only needs the interior camera
parameters. While classical stereo calibration needs the precise
geometric pose of two cameras, this method only needs images of the two
cameras.
- The benefit of HALCON 7.1: Uncalibrated stereo is revolutionary for
robotics, surveillance and security, where cameras are positioned at
widely differing positions, and thus, traditional calibrations are
quite cumbersome.
3D Position of Circles
Normally, determining the 3D position of objects requires 2 cameras. By
extracting and measuring circles (e.g., drill holes) in workpieces,
their 3D position can be determined with only one camera.
- The benefit of HALCON 7.1: Provided that the diameter and the camera
parameters are known, HALCON 7.1 can determine the 3D orientation based
on the geometric distortion from only one image.
64 Bit Technology under Windows
The new HALCON 7.1 version runs natively under Microsoft Windows XP
Professional x64 Edition. Consequently, the large Windows user
community can enjoy the large speed-up that Linux have known for quite
some time.
- The benefit of HALCON 7.1: Not only the underlying machine vision
algorithms profit from faster execution by using 64 bit processor
registers: With Windows XP x64 Edition now also machine vision
solutions that require more than 2 GB memory per process can be
realized. Compared to the 32 bit version, the computing time is reduced
by up to 30%.
AOP (Automatic Operator Parallelization)
Driven by the high amount of data throughput of every machine vision
application, the possibility to split the processing into several
threads that run on different processors has been used for a long time.
This is called multi-threading. HALCON automates the data splitting to
multiple processors (Parallel HALCON), a task that the programmer has
to solve with other systems. AOP was already supported by HALCON 6.0.
Nevertheless, this feature is of great use with the brand-new
generation of multi core CPUs now coming into market.
- The benefit of HALCON 7.1: After splitting the input data, AOP
distributes the data to the available CPUs to process them.
Subsequently, the computed data are merged automatically to achieve the
same results that are obtained with a single processor. All this is
done without any extra work by the programmer. In fact, this automation
yields a performance comparable to manually programmed multi-threading,
but at the same time brings many benefits: It is much easier to use, it
is not error-prone, and also programmers with little multi-threading
experience can immediately use it. MVTec Software GmbH
Diese E-Mail-Adresse ist gegen Spambots geschützt! Sie müssen JavaScript aktivieren, damit Sie sie sehen können.
www.mvtec.com
|