Development Of3-D Measurement System Based On Stereo Vision

With the progress of science and technology, sensor technology, computertechnology and image processing technology develop rapidly. The machine visiontechnology has been more and more widely used, and the application of machinevision technology in the non-contact measurement has become a hot researchproject. Researching non-contact measurement of3D geometry with digital imageprocessing and machine vision technology is significant to the theoretical researchand practical applications in product quality inspection, machining operation,precision measurement and robot navigation.Vision is one of the most important means for people to get information from theoutside world. Through the vision, human percept the size, brightness, color,movement of the object, and obtain various information significant to the survivalof the organism. At least80%of information is obtained through vision. As a result,vision is the most important faculty of human and animal. Machine vision is domeasurement and judgment with robots to replace human eyes. Machine visionsystem means to convert the measured target into image signal through machinevision products and send it to a dedicated image processing system, According tothe distribution of pixel brightness, color and other information, the image signal isconverted into a digital signal; the image system does various operations on thesesignals to extract the features of the target, and then control the action of theequipment according to the target results.(1)Through the vehicle images acquisition, this paper introduces the technologyof2D image processing in stereo vision with a single CCD image, includingbinaryzation, smoothing filter, sharpening technology and corner, edge detection.Therefore, image features can be recognized that were obtained by two CCD(2)Discussed how to realize3D image information in detail, including:three-dimensional imaging system, feature matching scheme, and theself-calibration algorithm based on Kruppa equation. On the basis, this paperresearched vehicle safety distance measurement method based on stereo vision.(3)Taking DM642as the core processor, this paper designed the hardware and software platform of the measurement system; and designed a simulationexperiment, which measureed and verificed traveling vehicle safety distance.