Study On Vision Measurement Technology Of High Temperature Forgings

With the development of computer vision technology, vision measurementtechnology has been used more and more widely. Due to high speed, high measuringresolution, non-contact and strong adaptive ability of vision measurement, it has beenthe hotspot of the research of non-contact measurement technology of hightemperature forgings. Based on the principle of computer vision, relative methods ofvision measurement technology of hot forgings are studied.High temperature forgings are different from other objects of normal temperaturedue to the heat and strong light radiated from themselves, which makes manyconventional measurement methods can not be adopted. Whereas, visionmeasurement can be employed in the field of dimensional measurement of forgings asa result of its unique characteristics mensioned above. Improvement of image quality,calibration of equivalent size of pixels, sub-pixel edge detection and cameracalibration etc. is completed in the paper. The main research contents are shown asfollows:1. The current status of online non-contact measurement of hot forgings at homeand broad is summarized and analyzed in the paper, so is that of vision measurementtechnology. Based on the above, the research significance of the paper is determinedand the technology routine is proposed.2. According to the theory of thermal radiation, the characteristics of radiation ofthe forgings in visible light range is discussed and the spectral response of CCD isanalyzed to derive the characteristics of the forging image, which paves the way forthe study of improvement of the forging image quality later.3. Based on the definition of image contrast, the optimization method of imagecontrast of hot parts is studied in the paper. First, digital optical filter is used toattenuate the R, G or B component of the image to different levels to improve theimage quality in terms of image processing. Then, in order to improve the imagequality in terms of physical hardware, different color filters are selected to install infront of the lens to control the light entering the cameras.4. On the principle of comparison measurement method, color camera and monocular camera are used to establish the two-dimensional measuring platform offorgings respectively. And then image denoising, calibration of equivalent size ofpixels, distortion correction and sub-pixel edge detection etc. is conducted to completethe measurement assignment. As being verified by the standard gauge blocks, themeasuring uncertainty in horizontal and vertical direction can meet the requirements.5. Based on the principle of binocular measuring, the symmetrical binocularsystem is set up in the paper. Through the mathematic model of the system, methodsof improving measuring accuracy are analysed and calibration based on the planarboard is used to derive the inner and outer parameters of the system. Before featurematching, median filtering, histogram equalization and image sharpening as thepretreatment procedure is applied to improve the matching precision. The featurepoints of sub-pixel level are exacted and calculated to get the3D world coordinatesand then to obtain the3D dimension of the forging.