Machine vision systems have measurement capabilities, enabling them to automatically measure the external dimensions of products, such as their outline, aperture, height, and area.
Dimensional measurement is an essential step in both the production process and the quality inspection after the product is manufactured, and machine vision has unique technical advantages in dimensional measurement.
Machine vision measurement principle
Industrial vision measurement technology (or digital near-field photogrammetry) is a stereo vision measurement technology. Its measurement system has a simple structure, is easy to move, has fast and convenient data acquisition, is easy to operate, has low measurement cost, and has the potential for online and real-time three-dimensional measurement. It is especially suitable for the detection of three-dimensional spatial points, dimensions, or the contours of large workpieces.
This non-contact measurement method avoids damage to the object being measured and is suitable for situations where the object cannot be touched, such as high temperature, high pressure, fluid, or hazardous environments. At the same time, the machine vision system can measure multiple dimensions simultaneously, enabling rapid completion of measurement work and making it suitable for online measurement. Furthermore, the measurement of minute dimensions is a strength of the machine vision system, as it can use high-magnification lenses to magnify the object being measured, achieving measurement accuracy of micrometers or higher.
For product dimension measurement, including one-dimensional, two-dimensional, and three-dimensional measurements, machine vision measurement methods are not only fast, non-contact, and easily automated, but also highly accurate. Among these, the measurement method combining a CCD camera and a microscope can perform minute dimension measurements, such as wafer measurement and chip measurement.
Measurement Principle: A CCD camera can acquire a two-dimensional image of a three-dimensional object, enabling perspective transformation between the actual spatial coordinate system and the camera's planar coordinate system. By combining two (or more) two-dimensional images captured by multiple cameras from different directions, the three-dimensional curved surface contour or three-dimensional spatial point positions and dimensions of the object can be measured.
Currently, the highest precision achievable using machine vision measurement technology has reached the sub-micron level, which can meet the precision requirements of most automated production at present. Measurement and positioning through machine vision systems can make production lines faster and more efficient.
Advantages of machine vision measurement
1. Machine vision measurement employs an advanced sub-pixel level object surface scanning method to meet the requirements of high-quality point cloud scanning. The system uses a high-resolution digital industrial camera to acquire image data. By observing the stripes on the object's surface caused by a light source, it can obtain a dense point cloud of any complex surface within seconds (the specific density depends on the size of the object being measured, the camera resolution, and the measurement distance; generally, the point spacing is 0.05–0.5 mm). The system uses resolutions ranging from 1.3 million to 5 million pixels to meet the needs of different customers.
2. A machine vision-based method for reconstructing the surface of a true-color object: The vision measurement system employs advanced image texture analysis and acquisition technology to maintain the true-color display of the object's surface while reconstructing 3D data. This technology effectively preserves the original color of the measured object, maximizing the reproduction of its true physical characteristics.
3. Fully Automated Stitching Method for Machine Vision Measurement. Image data from different perspectives are automatically stitched together in a unified coordinate system based on the texture of the object itself, thereby obtaining overall 3D image scan data. When scanning objects with rich textures, the system can complete the stitching function without pasting any reference points on the object surface, greatly improving stitching efficiency.
4. The measurement system's accuracy has minimal dependence on hardware. The entire system's optical calibration module uses ultra-high precision semiconductor technology, maximizing calibration accuracy. The software performs real-time error correction during data acquisition and rigorously corrects various lens distortions.
5. The machine vision measurement system is simple to set up and easy to use. Throughout the overall development of the 3D scanner, the principle of "what the software can handle, the user should not have to handle" was adhered to, minimizing the number of user-configurable parameters. During the scanning software's runtime, the required parameter values are dynamically calculated in a more accurate manner. This not only avoids unnecessary manual setup work by the user but also broadens the system's adaptability, increases its automation level, and reduces the possibility of human error.
Disclaimer: This article is a reprint. If there are any copyright issues, please contact us promptly for deletion (QQ: 2737591964). We apologize for any inconvenience.
