With the progress of industrial automation step by step, machine vision is no longer a strange term for most people. Machine vision has become the choice of employment for most students. Companies in the industrial field are also actively introducing machine vision talents and arranging industrial automation.
Computer vision and Machine vision
Some people are confused about computer vision and machine vision. In a broad sense, machine vision belongs to one kind of computer vision. Machine vision is the application of computer vision in the industrial field. But in a narrow sense, machine vision and computer vision belong to an independent branch. Computer vision laterally puts forward the characteristics of things from a large number of the same objects, so that computers can make rational judgments about external things like people. Pictures are actually a large number of digital information. When the algorithm can be used to read out the features of things in a large number of digital images, the computer can be more and more close to human judgment. This is what computer vision does, such as fingerprint recognition and face recognition.
Machine vision is different from computer vision. Machine vision often makes real-time judgment on the object to be measured on the industrial production line to determine whether it conforms to the specifications and whether there are defects. It does not need to detect a large number of production samples beforehand. It often uses the processing results of machine vision to determine the next step of implementation. Conduct guidance. The environment has a great influence on machine vision, because machine vision needs real-time acquisition of the image of the object to be measured, while computer vision focuses on image processing, and the requirements for image acquisition are very low.
Machine vision composition
As shown in the figure above, machine vision has the following components: light source, camera, lens, image acquisition card, software, control mechanism.
Light source, lens, camera -- machine vision's "eyes"
Many people do not understand the principle of machine vision, but if we compare the principle of human eyes to that of machine vision, we can find that the principle is the same. The light source, lens and camera are the eyes of machine vision. The light source makes the image features we need to acquire clearer. The lens focuses the light and reflects it on the optical sensor of the camera.
Image acquisition card -- "eyes" to the "brain" of the "nerve"
After the "eyes" acquire the characteristics of the object to be measured, they need to transmit information to the brain to make judgments. Then the image acquisition card plays the role of transmission. If machine vision loses the "nerve", the judgment of the "brain" will be terminated at this stage.
Software -- the "brain" of machine vision
Brain is used for thinking, and plays a very important role in the whole body composition. The brain makes judgments through the signals obtained by the senses, and then guides the movement of the body. For the software of machine vision, it processes and judges the acquired image features, so as to guide the next operation of the machine. So, software is the core component of the whole machine vision system. Do you think the brain can be unimportant?
Control structure -- "limbs" of machine vision
The limb is used for command and control, and the function of command is the software of machine vision. In the actual industrial scene, the software can continuously provide signals to the machine after running, so as to direct its next operation.
In fact, we have carefully studied and found that for machine vision, its replacement is the role of human beings. Machine vision is constantly moving in this direction. In future factories, we may seldom find human figures, because they have been replaced by machine vision, but machine vision still has a long way to go, after all, the program. In the future, the field of machine vision will continue to optimize and improve. It is a long and complex process to make machine vision have human thinking.