machine vision camera

Automating Factories with Cameras

    Machine Vision for Factory Automation

    In the first half of 2017, 19,331 robots valued at about $1.031 billion were sold to North American manufacturers. This is the highest rate of sales ever recorded to begin a year (an increase of 33 percent over 2016). The machine vision component industry also dramatically increased. This included smart cameras ($183 million in sales), optics ($20 million in sales) and lighting components ($35 million in sales).

    Smart Cameras:

    The smart cameras used in industrial automation have computer circuits that enable them to extract necessary data like scaling and important individual details from the image that can be directly applied to metrics. The cameras contain all the necessary communication interfaces as well as industry-proof 24 v input/output (I/O) lines for connection to programmable logic controllers (PLCs), actuator motors, and/or pneumatic controls. Smart cameras can evoke required response from the robotics in response to the condition of the manufacturing process or the output.

    In the past, robots were limited to performing repetitive tasks in setting that were very regular and consistent. They could be programmed to perform repetitive tasks very rapidly unless conditions changed. Today, the addition of machine vision through the use of smart cameras has enabled robots to perform more flexibly and effectively even where smaller production volume is required.

     

    Camera System Configuration:

    The configuration of vision systems is customized for various applications, to provide the robots with ongoing programming to cope with different visual conditions. Configurations of cameras in the production process may employ single cameras, dual or multiple cameras, patterns of cameras, time-of-flight sensors for laser measurement of distances, and 3D camera systems. The exact layout of the machine vision equipment must be carefully worked out when the manufacturing issues are planned. The output data from the cameras can be arranged into a logic that makes decisions about the way the motor parts of the system–the robots– function.

    Manufacturing engineers have to consider the kinds of lighting that will work best with the image sensors to get the best data from the cameras. On or off axis LED illumination, dome lighting, rear illumination or structured illumination may be required. The balance of illumination is a complex decision that may require considerable experimentation.

    Basic and Stereo Systems:

    Many “pick and place” applications require the use of only one camera. In the simplest case, the camera is placed in a fixed position above the object to be inspected and analyzed using off-the-shelf or custom software. For example, the robot placed near the camera responds to visual cues sent by the camera and picks the correct object placing it in the correct bin. The camera provides the robot with the location coordinates of the part and the robot locates the object and the bin.

    Single camera systems are used to pick out defective parts as well. Special lighting and filters on the lens enable the vision system to detect and respond to surface imperfections and instruct a robot to discard the defective part. To get a 360 degree map of the object being inspected, the robot can rotate the object inside the field of vision of the camera.

    One of the most common methods of determining the position of an object in the field of view of the camera is to use stereo vision smart cameras. The camera extracts information from two slightly different perspectives and computes depth from the disparity among objects in the images. Some applications require robots to know the exact position of objects. Cameras used in these sophisticated applications must be carefully selected for the proper resolution, focal length, and lighting. Helpful online calculators can help planners make decisions about the parameters of camera and camera placement in solving these kinds of problems. Often, the solution uses two separate cameras instead of one to build the stereo system.

    Integro Technologies can design and develop automated vision guided robotic inspection systems for a wide array of industries, and can help you with your automation needs. Our engineers can develop the perfect system to save you money and improve your production line thereby providing quality and consistency.