Application and working principle of visual sensing technology

The visual sensing technology is one of the seven categories of sensing technology. The visual sensor refers to the image processing of the image captured by the camera to calculate the feature quantity (area, center of gravity, length, position, etc.) of the object, and output. Data and sensors that determine the results. Vision sensors are a direct source of information for the entire machine vision system, consisting primarily of one or two graphic sensors, sometimes with light projectors and other ancillary equipment. The main function of the vision sensor is to obtain enough of the original image to be processed by the machine vision system.


How vision sensing works

Vision is a way of obtaining information about external environment from the biological world. It is the most effective means for natural organisms to obtain information and is one of the core components of bio-intelligence. 80% of human information is based on visual acquisition. Based on this inspiration, researchers began to install "eyes" for the machine, so that the machine can "see" the outside world information like human beings, thus creating a new subject - computer vision People have imitated the production of machine vision systems through the study of biological vision systems. Although this is quite different from the human visual system, this is a breakthrough in sensor technology. The essence of vision sensor technology is image processing technology, which is drawn into the image by intercepting the signal on the surface of the object and present it in front of the researchers.

Vision sensors have thousands of pixels that capture light from a single image. The clarity and detail of the image is usually measured in resolution and is expressed in number of pixels. After capturing the image, the vision sensor compares it to the reference image stored in memory for analysis. For example, if the vision sensor is set to identify a machine component that is correctly inserted with eight bolts, the sensor knows that the component with only seven bolts, or the component with the bolt misalignment, should be rejected. In addition, no matter where the machine component is located in the field of view, whether the component is rotated within 360 degrees, the visual sensor can make judgments. The appearance of visual sensing technology solves other sensors due to site size limitation or huge detection equipment. The problem of inoperability is widely welcomed by the industrial manufacturing community.

Vision sensing technology includes 3D vision sensing technology, 3D vision sensor has a wide range of uses, such as multimedia mobile phones, webcam, digital cameras, robot vision navigation, car security systems, biomedical pixel analysis, human-machine interface, virtual reality, surveillance Industrial inspection, wireless remote sensing, microscope technology, astronomical observation, marine autonomous navigation, scientific instruments, etc. These different applications are based on 3D visual image sensor technology. In particular, 3D imaging technology has an urgent application in industrial control and automotive autonomous navigation.

Intelligent vision sensing technology is also a kind of visual sensing technology. Intelligent vision sensor under intelligent vision sensing technology is also called smart camera. It is the fastest growing new technology in the field of machine vision in recent years. The smart camera is a small machine vision system that combines image acquisition, image processing and information transfer functions. It is an embedded computer vision system. It integrates image sensors, digital processors, communication modules and other peripherals into a single camera. This integrated design reduces system complexity and increases reliability. At the same time, the system size has been greatly reduced, broadening the application field of vision technology.

The intelligent visual sensor is easy to learn, easy to use, easy to maintain, easy to install, and can build a reliable and effective visual inspection system in a short period of time, which makes this technology develop rapidly. The image acquisition unit of the vision sensor is mainly composed of a CCD/CMOS camera, an optical system, an illumination system and an image acquisition card, and converts the optical image into a digital image and transmits it to the image processing unit.

Application of visual sensing technology

First, the car body visual inspection system

Body molding is one of the key processes in automobile manufacturing. The requirements for the body are strictly required, and 100% inspection of the body is required. The traditional body detection method utilizes a coordinate measuring machine, which is complicated in operation, slow in speed, long in construction period, and can only be subjected to sampling inspection. Usually, the key dimensions of the body are mainly the size of the windshield, the edge position of the door installation, the position of the positioning hole, and the like. Therefore, the visual sensor is distributed near these positions, and the spatial positional dimensions of the corresponding edges, holes, and surfaces are measured. The measuring station is designed on the production line. After the body is positioned, it is placed in a frame. The frame is composed of metal columns and rods distributed vertically and horizontally. The visual sensor can be flexibly mounted on the frame as needed. A corresponding number of visual sensors can be installed according to the number of measuring points (usually each visual sensor measures one measured point), and various types of sensors include binocular stereo vision sensors, contour sensors and the like according to different forms.

The working process of the measuring system is: the vehicle is transported from the production line to the measuring station for accurate positioning, then the sensor starts working in the required order, the computer collects the image of the detection point and processes it, calculates the spatial three-dimensional coordinates of the measured point, and calculates the calculated value and the standard value. The comparison results in the test results and the body is sent out of the measuring station.

Second, the cross-sectional dimension online vision measurement system

In industrial production, seamless steel pipe is an important industrial product, and its quality parameters are important data for manufacturing. The straightness and cross-sectional area of ​​steel pipe are the main geometric parameters, which are the quality of seamless steel pipe manufacturing. The key, but the measurement of the parameters becomes a problem for the following reasons:

1. The seamless steel pipe adopts non-contact measurement, and the manufacturing environment is bad;

2. The space size of the seamless steel pipe is large, which also requires the detection system to have a large measurement space. The emergence of visual sensing technology solves the above problems. The visual sensing technology uses non-contact measurement and has a large measurement range.

The measuring system is composed of a plurality of structured light sensors, and the light plane projected by the structured light projector on the sensor intersects with the steel pipe to be tested to obtain a partial arc on the circumference of the steel pipe section, and the sensor measures the position of the partial arc in the space. Each sensor in the system realizes the measurement of a partial arc on a section. Through appropriate mathematical methods, the spatial position of the section size and the center of the section is obtained by arc fitting. The spatial envelope of the center of the section is obtained, and the straightness parameter is obtained. Under the control of the computer, the measurement system can complete the measurement in a few seconds, meeting the real-time requirements.

Third, three-dimensional visual measurement

Digital measurement technology in 3D topography is the basic supporting technology for reverse engineering and product digital design, management and manufacturing. Its mechanism for digital measurement of 3D topography combines visual non-contact, fast measurement and the latest high-resolution digital imaging technology. Since the measured objects are mostly large objects with complex surfaces, the measurement is usually divided into two parts: local three-dimensional information acquisition and overall splicing. Firstly, the visual scanning sensor is used to measure each local area of ​​the measured shape, and then the splicing technique is adopted. The splicing of each part of the topography results in a complete image.

The sensor's visual scanning probe is designed using the principle of local binocular stereo vision. The overall splicing of the shape is essentially to put the collected data on the common coordinates, so that the overall data description can be obtained. The high-resolution digital camera collects the measured data from different angles and positions above the measurement space, and uses the beam-directional intersection adjustment principle to obtain the control point space coordinates and establish a global coordinate system. Finally, the correlation is performed through each coordinate system. Convert and complete data stitching.

How to choose a vision sensor?

At present, how to choose machine vision sensors in contemporary applications is more and more widely, how to choose machine vision sensors is worth learning, and now we have a deep understanding of how to choose machine vision sensors. The camera is the eye of the machine vision system, and the heart of the camera is the image sensor. The choice of sensor depends on accuracy, output, sensitivity, cost of the machine vision system, and a thorough understanding of the application requirements. A basic understanding of the sensor's primary performance can help developers quickly narrow down their search and find the right sensor.

Most users of machine vision systems recognize that the camera is a key element of the system and often treat it as a "chip" of the vision system. The camera itself is a complex system: including the lens, signal processor, communication interface, and the core part - the device that converts photons into electrons: the image sensor. The lens and other components work together to support the camera's functionality, and the sensor ultimately determines the camera's highest performance.

Much of the discussion in the industry has focused on processing techniques, as well as CMOS and CCD sensors. Both technologies have their strengths and weaknesses, and the processed sensors have different properties. What the end user cares about is not how the sensor is "how" is created, but how it performs in the final application.

In a given application, three key elements determine the choice of sensor: dynamic range, speed, and responsiveness. The dynamic range determines the quality of the image that the system can capture, also known as the ability to embody the details. The speed of the sensor refers to how many images the sensor can produce per second and the amount of image the system can receive. Responsivity refers to the efficiency with which a sensor converts photons into electrons, which determines the brightness level at which the system needs to capture useful images. The technology and design of the sensor together determine the above characteristics, so system developers must have their own metrics when selecting sensors. A detailed study of these features will help to make a correct judgment.