Advanced Materials Research Vols. 139-141

Abstract: Micro-displacement measurement with high accuracy plays a vital role in micro/nano manufacturing technology. In this paper, three low-cost techniques for micro-displacement measurement with high accuracy are discussed, and they are the methods based on DVD optical head, burred images and stop-action images. The techniques based on the DVD optical head and burred images are very attractive for many applications where the low cost is indispensable, and the method based on stop-action images has also better performance when measuring the static micro-displacement. Then, an application based on the stop-action images approach is given, and this method can measure the micro-displacement with nanometer accuracy. In addition, these discussions show that the methods based on the DVD optical head and burred images are a possible direction for low-cost micro-displacement measurement with high accuracy.

Abstract: A New Stereo Matching Method Based Phase Dynamic Programming (PDP) is presented in this paper. This method, based on gray code value and phase value, implements stereo matching through finding an optimal path on a 2D plane composed of cross phase absolute difference in the same gray region. Before matching, mask-image is adopted to reduce matching data point. During matching, the matching strategy is optimized by adopting sequence consistency technique as well as defining the path orientation. After matching, the error data point is eliminated by matching consistency and median filtering based disparity. Finally, experimental results show this algorithm can get a better performance. Matching process of PDP is simple but effective than others. It not only can get continuous and dense parallax but also eliminate effects of noise and the reconstructive object has high reliability, better authenticity and high precision.

Abstract: A good feature extraction method can improve the performance of pattern recognition system or classification system. Using potential energy theory into binary image feature extraction and feature store is a new method for image processing. The skeleton can be better display the whole features of the object. In target recognition system, using potential energy of skeleton-point projection into the plane coordinate system. The method can be better to show a skeleton in the structural feature. In addition, it can better avoid the matrix storage redundancy. In all energy projection method, potential energy projection is better shown its superiority in the structure information, the time of consumption and the storage space. The skeleton potential energy can be used in target recognition and target classification field and so on.

Abstract: As a tool of complex shape, the manufacturing error of　hobs can be classed into three kinds: the datum level error, the tooth thickness error and the parameter error of helicoidal surface. The parameter error of helicoidal surface, which is composed of tooth error and helix error, is the most importance among them. For precision of hob and convenience of measurement, several individual errors are instead of the helix error. How to set reasonable individual errors to guarantee the helix error within tolerance, the relationship between them in gear hob measurements is analyzed, and the approach to decompose the helix error into individual error has been discussed in the paper. On the basis of above theory, a high precision mode can be easily established using the soft ware.

Abstract: Detecting the weld position is the precondition of seam tracking during an arc welding process, and is crucial to achieve high-quality welds. Visual sensing is one of the most attractive approaches to detect the weld position, which provides valuable information to control the arc following the expected path. The weld edge is main characteristics of the weld position image, here the visual technique for detecting the weld position based on Hough transformation is proposed. The weld position can be detected and recognized through Hough transformation, while the weld edge image is detected using the Canny operators. And Hough transformation is modified based on the actual weld situation. Therefore, the precision of the weld detection can be improved. From the analysis of experiments, it has been proved that the proposed method can detect the width and the center of the weld, and is suitable for the seam tracking system based on the computer vision.

Abstract: Improved Geometrical Product Specifications (GPS) is an internationally accepted standards system. It is the foundation of the technology standards and metrology specifications. Uncertainty is one of the most important concepts in improved GPS. In the improved GPS standards system, uncertainty is used as an economic tool to enable optimum allocation of resources amongst specification, manufacturing and verification. Estimation of uncertainty in measurement according to improved GPS can improve the reliability of the measurement. A method to estimate the uncertainty in straightness measurement is proposed according to the requirements of improved GPS standard system. Based on the mini-zone method and the transparent box model, the calculation equation of uncertainty in straightness measurement is deduced. The experiment indicates that this method may assure the integrity of the verification result and improve the reliability of measurement.

Abstract: In this paper, a surgical blade inspecting system is designed and researched, which can automatic test and determine eligibility according to tolerances. High precision and non-contact detection of surgical blades based on machine vision is realized by using camera with telecentric lens to obtain the image of surgical blades, by applying template matching technique to automatic match surgical blades. One-dimensional edge of sub-pixel accuracy is extracted by combining the edge filtering of the Deriche filter and parabolic fitting for the maximum range from edge. It can be used for precise measurement of different part of surgical blades with different sizes. Extensive experiments showed that measuring range is from 2mm ~ 50mm, measuring accuracy is 10μm and average detection time is 0.6s. Thus the inspection system is fast, accurate and robust to fulfill the industrial demands.

Abstract: The traditional method of wire detection is implemented artificially, and that is heavy-worked, inefficient and low-accurate. This paper proposes a novel method to complete the job by using image technology to improve production efficiency and product qualification rate. In this paper, Pulse Coupled Neuron Network (PCNN) has been applied to the wire image preprocessing, and then Hough Transform (HT) is used to extract features of lines, and detect them according to these features from the processed image. After that, whether the wires were qualified was determined by the criterion of parallelism and isometry. Comparison between different wire images was given in the experimental study. Experimental results showed that the proposed method was feasible and this method could achieve higher detection accuracy and efficiency compared with the traditional method.

Abstract: Image measurement based on machine vision is a promising method for the precise measurements of machine parts. As for a large sized workpiece that exceeds the FOV (field of view) of the camera, image mosaic must be performed to implement measurement. While traditional template matching algorithms will lead to long runtime consumption, a combination of algorithms with precise mechanical positioning function using motion control was proposed and applied in the machine vision system. The manufacturing and installing errors have been taken into account to improve the positioning accuracy. This method minimized the search space and thereby reduced the runtime substantially. In addition, a stop criterion of the computation for the algorithms was applied to save a considerable amount of time in the process. Experimental results show that the precision of the image mosaic meets the requirements while speeding up the process significantly.

Abstract: In plate rolling, the existence of some asymmetry rolling conditions, affects the distribution of transversal thickness of rolling pieces, leading to the occurrence of camber phenomenon, which in turn seriously affect the product's material yield. In this paper, based on the images of plates in the production line collected by camera, taking into account the requirements for on-line measurement, filtering noise from image information, use edge detection operator, Hough transform methods and sub pixel edge location method to identify the steel plates, have achieved in-line measurement of camber curvature of plate side. At the same time, combined with camber control model, we have developed camber curvature measurement and control system, realized the automatic correction control of camber curvature of rolling pieces in the production process. Field applications show that this method can reduce the human intervention, will limit the camber curvature of rolling pieces within a certain range, which has great significance for improving the product’s yield.