| Large components are the key parts of important equipment which are widely used in the field of power, ship and aerospace. In order to let the large components can be employed in extreme condition, the blanks which are used to machine the large components are manufactured through forging. The large size forgings have to be measured in the forming process. However, the high temperature of hot large forgings and the large size make it hard to be measured accurately. Machine vision technology has been used as a high-speed, non-contact method in dimension measurement. In order to reduce the waste and production cost, the dimension measurement method with the machine vision technology is researched. This paper is mainly focuses on some key technical as follows:(1) To solve the problem of acquiring the image of hot parts, the spectrum selective method is proposed. By analyzing the self-emitted radiation character and reflection character of hot large forging, the illumination lamp is used to acquiring image of hot large forging. When the external light source projects light toward the high temperature objects, the external light strikes the hot objects and is reflected toward CCD cameras along with the self-emitted radiation. One component of the reflected light is selectively detected by the CCD cameras, and the image of hot objects can be acquired. The range of wavelength is obtained according to the wavelength when difference between the radiant exitance of illumination lamp and the radiant exitance of hot forging is the maximum value. This method presents an important base for measuring the dimension of hot forging.(2) Considering to the measurement condition in the forging plant, a stereo vision system which using the spectrum selective method is developed for mesuring the dimension of hot large forging. Structure light is projected toward the surface of hot forging in the dimension measurement process for increasing characteristics on it. The effect of the structural parameters of the stereo vision system on the error of the measurement is analysised. Based on the analysis results, the distance of two cameras, the focus length of cameras and the distance between the stereo vision system and the hot forging are adjusted accordingly.(3) The camera calibration technique that can be used to calibrate the dimension measurement system in the forging plant is proposed. The intrinsic parameters of camera are calibrated according to the active vision. By recognizing the corresponding points in the images acquired after each orthogonal translation, the intrinsic parameters can be calculated. The extrinsic parameters of camera are determined up to indeterminate scale by the essential matrix. Then, scale is calculated by dividing the distance of two reconstructed feature points in 3D space by the length of the two feature points which is measured by the double-theodolite system.(4) In order to measure the dimension of hot forging in the forming process, the method based on projected feature lines which can obtain the shape of forgings is presented. When measuring the dimension of round forging, a coded single pattern is projected toward hot forging to label characteristic lines and simplify the process of finding correspondences between image pairs. If the feature line is vertically to the axis of forging, the diameter of round forging can be estimated by fitting an ellipse using the reconstructed feature line in 3D space. Otherwise, the diameter of round forgings are acquired by fitting a cylinder with the reconstructed feature lines in 3D space. The lengths of square forgings are measured by finding the position of the curvature which is greater than the threshold.The spectrum selective method for acquiring the image of hot part, the dimension measurement method by reconstructing the featured lines that are projected on the surface of hot forging, the measurment system presented in this paper can be used to measure the geometric parameters of hot large forging effeictively. |
PDF Full Text Request