Advances in machine vision systems

Manufacturers of vision system technologies are finding it extremely challenging to keep pace with developments in the semiconductor industry.

Miniaturisation of components is creating the need for higher throughput vision systems that offer superior accuracy levels.

"Commensurate advancements in lighting and illumination, lasers, lighting, processors, sensors, and optics have to complement machine vision systems," reiterates Frost and Sullivan research analyst Vishnu Sivadevan.

"Developers face the challenge of reducing set up time and also incorporating greater enhanced functionality, scalability, and upgradeability".

It is important to pay special attention to issues surrounding the expenditure and time on for set up time and installation.

User-friendly features, robust integration capabilities, and reduction of operator training time will drive investments in machine vision systems.

"End users are demanding quick return of investment and are attempting to make a futuristic investment in vision systems, choosing judiciously from a range of products, which differ in cost and functionality," explains Sivadevan.

Advanced automation standards also require machine vision systems that are flexible and scalable across multiple products and production lines.

This has given rise to compact vision systems and smart cameras that have built-in image sensors and processors and are more user friendly.

While smart cameras are replacing PC-based vision systems, the choice of architecture would depend upon the application to which the vision system is catering.

The increasing need to optimise production processes is one of the key drivers of development of innovative machine vision technologies.

Basler Vision Technologies, Germany, a leader in the manufacture of vision-based optical media inspection equipment has developed a laser-based measurement approach.

The third generation optical media - the blue ray disc requires highly accurate cover layer uniformity and has posed a significant challenge to machine vision system manufacturers in reducing cover layer thickness and the microscopic surface deformations.

Basler's scanners enhance the optical effect of deformation and can spot defects that are present either on the information or covering layer of the disc.

Once vision systems achieve greater versatility and upgradeability, the range of applications is likely to expand beyond industries and production processes.

High-end machine vision applications are progressing from two-dimensional (2D) to three-dimensional (3D) imaging with techniques such as laser triangulation and stereovision.

3D chip-based vision systems are emerging as cost-effective alternatives to weight sensors and stereovision techniques in automotive applications.

"Upgrading to 3D inspection systems from 2D inspection systems would constitute a phenomenal leap in performance for certain applications," observes Sivadevan.

Sick IVP, Sweden has introduced a first of its kind 3D vision smart camera that uses laser triangulation for high-performance capture of 3D images.

Due to 3D inspection capabilities, the camera has versatile qualities that enable it to perform inspection, location, and measurements of objects to enhance production processes.

This revolutionary 3D smart vision system, which also incorporates tools that can estimate height and volume is set to make a significant impact on robotic guidance applications such as bin picking, inspection of connector pins, and inspection of weld seams all of which require precise inspection of 3D images.

Currently, researchers are also working toward the development of real-time autonomous robotic guidance using machine vision systems.

The five year reverse engineering the vertebrate brain (Reverb) project jointly undertaken by the Engineering and Physical Sciences Research Council (EPSRC), BAE systems, and a group of universities in the United Kingdom aims at incorporating artificial intelligence in robotics.

As part of this project the department of electrical and electronic engineering at the University of Manchester has developed a vision chip, which is capable of foveal and peripheral vision similar to the retina of the human eye and is likely to be extremely useful for factory automation applications.

Termed as a 'smart sensor', this vision chip performs the functions of a vision sensor and a microprocessor is capable of processing complex images at rapid rates, and will find use in laser-guided crawlers for carrying out tasks such as machining and inspection of aircraft parts.