Computer vision is a mathematical algorithm of recognition of 3D objects on images and videos. Computer vision as a separate discipline started in 1960 as the study of artificial intelligence in universities. The first studies of computer vision were consisted of a camera connected to a computer and answering the question ‘describe, what you see’.
Right now computer vision impacts industry, consumer, scientific researches and other fields. Computer vision developed as a separate discipline with its own algorithms and analysis tools. Nowadays computer vision plays a big role in the following applications:
- 3D model building
- Optical character recognition
- Medical imaging
- Automotive safety
- Fingerprints recognition and biometrics
In this course we will mostly concentrate on the industrial use of computer vision. Systems based on computer vision can execute quality tests, manage robotic systems, perform control tests, identify parts and read valuable information, which makes them a great asset in manufacturing. Machine vision is an important part of many robots in manufacturing and likewise there are similar advantages in computer vision:
- Costs savings
- Absolute product quality checks from sides that can not be seen with the human eye
- Processes optimising
- Flexibility and sustainability
- New technologies usage
The computer vision systems market is growing in industry. Future manufacturing will be controlled, supported and managed by robots, equipped with computer vision and artificial intelligence systems. These robots and systems will bring with them benefits in terms of low consumption of energy and materials. Production processes will also be more flexible.
Product quality checks are vital in manufacturing – non–compliance to specification and other problems and losses caused by random checking can be resolved by computer vision systems. Computer vision is relatively cheap in terms of the cost of inspection of all the manufacturing products. Quality control is evolving to support the optimisation of production at very early stages.
Visual inspection of the product at the various stages of production may provide notification if product parameters deviate from its norm, preventing failures of products at the final production stages.
The world of connected things is growing further and further – the Internet of Things (IoT). A large number of devices are exchanging data for operation, and here vision and analysis are an important, and often crcial piece of information. Computer vision is also an application for security systems – cameras for condition monitoring, tracking cameras and other devices. 3D cameras can also be used in rare application and in everyday life.
The most important features of computer vision that may improve the quality of a computer vision-based system, and as a result the product or service are; the 3-dimentional vision and analysis, which helps to bring inspection precision to the new level of quality and avoid product failures; standartisation – the common standards and protocols which allow the use of computer vision devices as a part of a bigger analysis and executing system; performance – the performance of the components for computer vision devices and systems is getting better and better, rapidly increasing performance of the full system.