Smart Cameras optimize quality control

Ettlingen – Intelligent Cameras execute increasingly demanding tasks in industrial image processing. The devices developed by Vision Components GmbH provide an impressive example: the Smart Cameras manage without an additional PC and can be used as stand-alone image processing systems. The following text showcases an application example in which a fully automated, high-precision quality control facility was implemented by means of three VC cameras.


Sintered metal moldings measuring 5 mm, which are used in automotive industry applications, are tested. The parts must be examined for correct external measurements as well as true running, axial runout, inner diameter and the occurrence of ridges, injections and cracks. The required overall true running tolerance for the inner diameter and the overall axial run-out tolerance for the flanks and the front sides is 50 µm. The quality check is rendered difficult mainly because of the multiple test dimensions. Additionally, the parts have no axis which could serve as a basis for the true running test.

Testing facility in detail

The testing facility was designed and implemented by Wente/Thiedig GmbH, a specialist in industrial image processing solutions, in collaboration with the company Wagner-Zuführtechnik. At first, the isolated moldings are centered by means of two pins. Then, they are put into rotation, creating a “virtual axis” which can be used for the true running test. The pressure exerted by the rounded tips of the pins is precisely adjusted so as not to damage the parts and not to distort the subsequent diameter test. The individual measurements of the true running test are recorded by a VC2068 camera from Vision Components. Telecentric illumination ensures undistorted images. (The VC20xx camera series, which was state of the art at the time of commissioning and is still being used in many applications, has meanwhile been replaced by the follow-up series VC40xx and VC44xx – cf. section “Smart Cameras for industrial applications”).

Figure 1: Molding component in the true running test

Subsequently, the molding is taken over by a grabber and positioned between two other VC2068 cameras which determine the inner diameter and check both flank exteriors for cracks. The calibration of the testing facility requires no throughput of master parts: the standard parts for the true running and the diameter test are integrated into the machine. In order to check the system, calibration intervals can be chosen freely.

Figure 2: Molding component in the inner diameter and crack test

Data acquisition and analysis

The parts are examined for true running by one camera using 30 measuring windows. Based on an adjustable number of individual measurements per run (with a maximum of 100), these windows enable the camera to determine whether the measurements are within the pre-set parameters. Approximately 1,100 boundary points are analyzed for the diameter test. The three cameras are linked via Ethernet and communicate with the PLC via their digital I/O channels. The results of the individual tests are transferred to the PLC, which analyzes the acquired data and rejects the moldings or sends them to a container for OK parts. A PC is connected for documentation and remote maintenance purposes.

The cameras operate at a measuring frequency of 72 ms. The overall cycle time of the system depends on the number of individual measurements per true running test. In this particular application case with 15 measurements, the system reaches a cycle time of 3.2 seconds. The testing facility provides high measuring accuracy: in the true running and the diameter test, the measurement uncertainty is less than 1 µm. It ensures reliable, precise and fast quality control – such extensive checks could not be realized by means of conventional mechanical methods.

Smart Cameras for industrial applications

Vision Components provides integrated image processing solutions for industrial environments which only require appropriate application software for use in specific applications. All Smart Camera models feature a compact housing suitable for industrial use. They integrate all required components for image processing (image sensor, processor, frame grabber, interfaces and several PLC-compatible I/Os).

Figure 3:
The real-time cameras are not susceptible to viruses. Their key benefits include virtual freedom from maintenance and low hardware expenses. Requiring neither hard drives nor fans, they are more reliable than PC stations. A powerful digital signal processor enables the Smart Cameras to execute all image processing routines. Thus, they can completely replace conventional PC-based systems. Depending on the type, the cameras have a maximum computing power of 8,000 MIPS – the VC44xx series, for instance, is based on 1 GHz Texas Instruments DSPs, the fastest models currently available on the market. Image recording, which can be triggered externally or internally, is carried out by a high-quality CCD sensor with a resolution that can be chosen between 640 x 480 and 1,600 x 1,200 pixels. Depending on the resolution, the cameras have a frame rate between 10 and 242 fps

Figure 4: High-performance camera VC44xx

All camera functions are controlled by the Linux-like real-time operating system VCRT. All camera functions can be operated via its integrated shell, which also allows users to execute file operations, start programs and recall TCP/IP statistics. The file system is protected against power failure; and shutting down simply requires switching the power off. The Smart Cameras can be freely programmed in C and C++, allowing users to adapt them to specific applications. Programs are developed in an integrated development environment (IDE) on a PC. The finished program is then transferred to the camera (e.g. via FTP) and tested there. Please visit the company’s website: