Search

Larger workpiece's size visual detection needs to combine motion control and image mosaic to ensure accuracy. Usually the complex flat parts have many test items therefore require tedious path planning which affect detection efficiency and accuracy. This paper researched and developed a intelligent path generation method using image interaction. The method is based on DXF drawing file analysis, combined with smart interaction, image restructure and sample point automatic positioning to achieve plane size detection automatic processing and path planning. Application results shown that this method surely reduced difficulty of complex workpiece's detection process planning and increased detection system's efficiency and accuracy. Furthermore, its easily extended application to 3D or other measure system also improved intelligent level.

On the basis of analyzing the blur images with noise, this paper presents a new segmentation method which is based on the morphology method, fuzzy K-means algorithm and some parts operator of the Canny algorithm. Because of the Canny's good performance on good detection, good localization and only one response to a single edge, we introduce the course of Canny operator that calculating the value and direction of grads, non-maxima suppression to the grad value and lag threshold process into our post-treatment process. Through experiments, it is demonstrated that the image segmentation method in this paper is very effective.

The paper studies the intelligent detection system based on TCP/IP protocol and realizes real-time detection based on B/S mode. This article focuses on the real-time problem of the communication system, an also propose an improved algorithm of congestion avoidance based on receiver-driven(CARD). Being different from the RTT algorithm and some traditional algorith,CARD uses one-way delay detecting at the receiving terminal to feedback congestion levels, and change sending rate in accordance with a linear function of congestion level dynamically, avoiding entering the congestion situation. Simulation results show that, CARD can avoid the congestion more efficiently and improve the real-time performance of the system .

The algorithm of fractal dimension and co-occurrence matrices is proposed and is applied to material Vickers hardness image segmentation. Based on the characteristics of the indentation images, this article uses texture features to extract the indentation silhouette from the point view of texture segmentation. We adopt fractal dimension and co-occurrence matrix algorithm to describe the texture characteristics of the indentation image, forming a n-dimensional feature vector, introducing EPNSQ to smooth the features. Finally we combine with the k-means clustering algorithm to get texture segmentation result. The experiment demonstrates that in the material Vickers hardness image segmentation the proposed algorithm was significantly effective and robust.

There are many vital defects in the process of fabrication of IC wafer. The redundancy material defect is an important factor of reducing the yield of IC, it mainly causes circuits to be connection failure, so it is important to recognize the redundancy material defect and ascertain the circuit's failure style. In this paper, a simple and doable method of vision inspection for the redundancy material defect is presented. According to the image character of taken images, in order to improve image quality, the method of light compensation and morphological filter is used to eliminate the disturbed noise. Then, the projection theorem and the inspection method based on character of pixels distribution are used to recognize the connection circuits.

In this paper, structure and working principles of eddy current retarder (ECR) are discussed, as well as performance requirements of real-time virtual testing and simulation. On the basis of hardware testing platform, virtual models of vehicle and retarder are embedded into application program. In the testing system, Vega API function library is used for second development to control the 3D models, and Visual C++ is used for mathematical models call from MATLAB to simulate the working condition. Integrated with virtual instrument and virtual environment, the limitation of former simulation is broken through. Monitoring and testing on eddy current retarder are achieved efficiently and visually. Besides, the virtual test data is not only compared with the practical vehicle test data to correct the technical parameter of eddy current retarder, but also compared with the simulation of mathematical models for further study. The testing system gives evaluation on the performance of the retarder in different vehicles, so it expands application range and heightens testing efficiency.

In this paper, based on the electrothermal upsetting processing method, the process of valve upsetting is divided into two steps: initial upsetting and formal upsetting. Four parameters, including 1.hearting current; 2.press; 3.velocity; 4.material, which influence the manufacture process in evidence, is mainly discussed. Analyzing defect and deficiency of tradition processing, the article puts forward to a controlling strategy with iteration technique and designed an embedded valve upset forging microcomputer control system. The electrothermal upsetting control system based on embedded technology stores the experience and technique of local worker, optimizes the tradition processing by using ARM S3C2410X chip. During the process, controlling parameters such as press, displacement and voltage are real-time monitored. Signal feeds back to system by sensors to realize the closed loop controlling on-line self-study function. And the technics parameters and the controlling process will display on the HMI (Human Machine Interface). Using the system, the quality and efficiency of the valve manufacture are improved, and the robustness gets stronger. The system has been applied in enterprises. Comparing with the tradition one, the automaticity in manufacture is improved, and the rejection rate is reduced. It also benefits the informationization of production of enterprise.

High SNR (signal to noise ratio) image is deeply needed in the precise analysis of IC wafer micro-image; current denoising algorithms cant reach the analytic precision in some level. In tins paper, one image denoising algorithms is putted forward based on adaptive lifting scheme, the construction of Haar wavelet and CDF (2,2) is given, the signal is decomposed by wavelet base Haar or wavelet base CDF (2,2) adaptively along four directions (horizon, verticality, 45 degree and 135 degree) in the step of predicting, the wavelet coefficients are calculated separately at each direction, all the thresholds are gained using wavelet soft-thresholding principle, the optimal thresholds minimize the error of the result as compared to these the signal is decomposed along horizon and verticality. Finally the definition of the image is appraised with gray gradient judging function, the experimental data shows that the focus error is no more 4 um, the display definition of the image is improved.