Hardware design of embedded machine vision measurement system for bearing RING diameter size
ABSTRACT
Vision measurement uses image processing algorithms, by calculating the number of pixels between the images corresponding to the object border to determine the object's geometrical diameter. This design applys machine vision and image processing technology, and also uses a CCD camera, a image acquisition card, a digital image processor and a display to form a circle embedded machine vision system for bearing ring diameter size. In the system, this design mainly designs the module of optical imaging, A / D conversion for analogy signals, image processing and communication.
This embedded machine vision measurement system works as follows : In the light source, bearings’ images are projected onto the optical imaging system, and enlarged by the lens, after that ,the images are focused on the CCD surface of photosensitive array, camera converted the optical images which has already been received into video signals which will be transmitted to a video decoder; and then, the video decoder will converted the analog signals into digital signals; then under the control of sampling controller of CPLD, FIFO will collect the digital image signals; the image processor in system mainly completes the data acquisition from FIFO and also fulfills the algorithm of images enhancement, denoising, edge extraction, the calibration to measurement system and progressive scan, by calculating the number of pixels between the images border corresponding to the measured bearing to determine the geometrical diameter of the bearing, finally results will be transmitted to the display.
Machine vision system owns the virtue of non-contact, on-line real-time, speed, high precision and the scene anti-jamming capabilities, and can achieve the production target of zero-waste products.
KEY WORDS:embedded, machine vision, bearing ring, measurement, DSP, TMS320C6201