It is difficult for a production system error-free or find perfection in the absence of computer vision, we are pleased to be informed when errors creep into the process so we can take corrective action, and therefore the quality and productivity are maintained. Systems equipped with extensive visual features are adaptable and can cope with environmental uncertainty. As the safety and reliability, product quality and flexible automation solution is the most common problems in the industrial environment, so the adaptation of a powerful vision set by the needs of this environment is a kind of complex set of ways to show results Zou system.

One of the most important issues in computer vision is an automated visual inspection, is a great need for help with this problem in areas such as quality assurance, product inspection, testing and calibration. These applications can be classified into measurement, inspection or sorting.

The investigation concerns the measurement of parts and the control of tolerance, in which the inspection process of the implementation of a partial examination (ie, missing parts or sections of the test object), and finally, the classification concerns the identification and recognition of objects or parts. Today, the vision of the handling robot for automated industrial environment less varied in two dimensions rather than the complexity of dealing with three dimensions and controls.

The only two applications are also developed processes for promoting and welding, is the reason why the applications are easily reduced to two dimensions. All systems of practical machine vision in use today are for their particular purposes. Some are used to ensure that the parts come out lines are correct. Some lines are used to provide a way for cars to carry themselves to recognize.