Every mechanical product is assembly of parts according to the parts' technique requirement. So the quality of product mainly decided by processing precision of parts and mechanical assembling accuracy. But the processing precision of part is decided by dimension, shape and position accuracy. Shape and position accuracy is one of the most important technique point to assess product quality.Straightness measurement is one of the most basic and important item in the field of the geometrical measurement, which is the base of planar degree, parallel degree, orthogonal degree, coaxial degree and etc. The straightness is also an important content in the field of mechanical manufacture. With the dimension accuracy, circularness and roughness, they are called four factors to affect product quality. But the instrument's precision is affected by part dimension, shape, position and etc .As one of the main factor of shape deviation, the straightness directly affect instrument's precision, function, quality. However in the factories, the straightness measurement's traditional way is old and low precision, so we take the measurement of 3-D straightness as the object of this thesis.In this thesis, the methods to measure the 3-D straightness and the result data assessment are discussed. After compare of several traditional ways, we use the ray axis approach based on the digital image processing, and design mechanical structure. Sensitivity and speed of measurement is improved by using the optical enlargement principle, beam splitter and CCD camera. At the same time, the way of image processing is discussed.the adopt image processing way to our measurement method. Using LabVIEW, software which can read images, remove noise of images, change gray images to two-value images, open and close operation is programmed. Also assess the data processing of 3-D straightness is assessed by use this software. Comparing the method of lining two points which are the termination of the data, the least square fitting method and genetic algorithm, genetic algorithm is selected to assess 3-D straightness. |