Weixing Zhu

A Fast and Accurate Algorithm for Chessboard Corner Detection

The authors point out the limitations of SUSAN corner detector in detecting chessboard corner, then describe an improved SUSAN(Smallest Univalue Segment Assimilating Nucleus) detector algorithm for detecting chessboard corner on the basis of symmetrical geometry structure of USAN (Univalue Segment Assimilating Nucleus) area. And the algorithm has been applied to the chessboard images on real photos. The improved algorithm can quickly detect corner from real photos shot from every angle. The theory of detecting corner at sub-pixel level is Orthogonal Vector Theory, that is, vector from the corner to its adjacent area pixel point should be vertical to gray grads of the adjacent area pixel point. In order to get the coordinate of corner at sub-pixel level, we establish the neighboring area equation and solve it via iterative method, and propose to check its validity according to cross ratio invariability in perspective projection. Keywords-insert chessboard corner; USAN; sub-pixel; cross ratio invariability

A New Search Algorithm Based on Muti-Octagon-Grid

To cope with the defects of the fast search algorithms based on the UESA (unimodal error surface assumption) that can easily lead to a local minimum or wrong search direction and on the analysis of the characteristics of many kinds of motion sequences, a new search algorithm based on the Muti-Octogon-Grid (MOGS) is proposed. The algorithm takes adaptive search strategies by using the distribution characteristics of motion vector and some other technologies, such as the adaptive threshold of macro-block and dynamic search range. Experimental results show that, compared with the UMHexagonS and Full Search, the algorithm maintains the same image quality and bit-rate, but it makes great improvement on the search speed and encoding time. The PSNR (Peak Signal-to- Noise Ratio) increases about 0.004dB, the SIR (the Speed Improvement Rate) is 42% and 91%, the encoding time can be reduced by 21% and 78%, only with the bit-rate increased by 2.1% and 3.0% respectively on average.

Research on quantitative analysis method of steel dimples based on wavelet transform

In order to explore the relation between steels dimple size and mechanical properties,it is necessary to measure and classify the dimple size accurately. An improved dimple image quantitative analysis method based on combination of wavelet transform and mathematics morphology was proposed in this paper. It measured and analyzed the dimples by means of edge detection based on combination of B-spline wavelet transform and mathematics morphology, water-growing segmentation and measurement method etc. Application results showed that it had a good processing effect to solve the defects such as dimple edge fuzzy, dimple edge deletion and dimple holes etc, and its quantitative analysis was accurate, effective and convenient.

Pig target extraction based on adaptive elliptic block and wavelet edge detection

In order to obtain the prospect targets from the top view image of pig barns, the target extraction method of pigs based on adaptive elliptical block and wavelet edge detection is proposed in this paper. Firstly, two-dimensional OTSU global threshold segmentation and mathematic morphology processing were used to get the initial segmentation results. And then, adaptive elliptical block were taken around each pig respectively in the original image according to the initial segmentation results. Finally, the wavelet edge detection for secondary accurate extraction was used in each elliptical block and the complete contours of each pig were obtained. This method is suitable for a complex environment of the light changes in pig barns. This work provides a good foundation for the further research of pigs individual identification and behavior analysis. On the other hand, it also explores a new way to extract target contours of other similar images in complex environments and scenes.

Design and application of Decision Support System for plastic die materials selection

Applications of Decision Support System for plastic die materials selection were discussed in this paper. Decision Support System was established by combined entropy method with TOPSIS method, among which entropy method was used to determine the attribute weights of the candidate materials and TOPSIS method was used to sort the candidate materials. Test results showed that the system could provide a reasonable project of materials selection for users.

Study on thermal-fatigue property prediction of hot-working steel based on PLS

An innovative approach based on partial least square is employed to predict thermal-fatigue property of hot-working steel. By extracting principal components, correlation information between properties of material and thermal-fatigue property under the operating temperature is screened and synthesized. Based on the research results, thermal-fatigue property prediction model of hot-working steel is established. The results show that the model can effectively predict thermal-fatigue property of steels. Additionally, the thermal-fatigue property can be improved by enhancing the material yield strength at high temperature, impacting toughness, thermal stability, oxidation resistance and elongation rate.

The Research on Measuring Analysis and Prediction of the Nanoscale Precipitated Phase in Steel

The grain size and shape of precipitates in alloyed steel were analyzed and predicted for promoting the refinement effect of nano-nucleation adequately and optimizing the thermal simulation technique. The mapping relationship between thermal simulation parameters (including chemical composition, deformation temperature, and deformation amount and holding time) and nano-precipitates state including grain size, shape and distribution was studied. The weight of BP neural network prediction model was improved by L-M algorithm. The shortage of conventional B-P algorithm such as the slow speed of training, easily come to a local minimum and weak of global search was overcome. By the means of the simulation and practice, the prediction precision of the improved BP neural network on the grain size of nano-precipitates was more than 93%, and the precision of the shape of nano-precipitates was more than 90%.

Ultra-Fine Grain Steel's Metallurgical Image Restoration Method Based on Improved Water-Growing

In order to solve the unique problems, such as grain boundaries fuzzed or disconnected and so on, in the ultra-fine grain steels metallurgical images, a metallurgical image restoration method based on improved water-growing is proposed in this paper. According to the irregular morphological characteristic of overlapped grains, the grain boundaries are obtained by ultimate erosion, numbering the objects and repeated dilation. So that the overlapped grains are separated and the images are restored. Compared with the manual restoration results, the experiment results indicate that this method can efficiently restore and reconstruct the defect grains and other defects in metallurgical images. Moreover it can effectively solve the problems associated with grain boundaries fuzzed or disconnected etc, and provide the solid foundation for quantitative metallurgical analysis.

Research on Measurement and Classification System Design and Application about Ultra Steel Dimple

In order to search the relationship of steel dimples diameter and mechanical properties, it is necessary to measure steel dimples diameter exactly and make a good numerical statistic, which can reduce the influence of human factor possibly. Adaptive fuzzy threshold segmentation method for equiaxed dimple image analysis and binary segmentation, which is produced by the ductile rupture, is used in this article, so the dimple and back ground are separated well. In order to measure exactly the dimples diameter, the algorithm of multi-region area based on freeman chain is improved, which can measure steel dimples diameter effectively, conveniently and accurately. Moreover it can provide the reliable foundation for analyzing the steel dimples diameter in the future. In this paper, an automatic analysis method on steel dimple is presented, which is based on adaptive fuzzy threshold segmentation and the improved algorithm of multi-region area. Therefore, this method about the dimple analysis is automatic 、efficient and precision.

Dead Birds Detection in Modern Chicken Farm Based on SVM

A major labor-intensive task still remains in modern chicken farm, that is timely detection and removal of dead birds. In order to solve this problem, an automated dead bird detection method based on SVM (Support Vector Machine) was proposed. Firstly, take two pictures sequentially on the same cage, then process the images and extract the contours of bird. Secondly, a center radiating vector representation is used to abstract features of the contour. Minus the corresponding vectors in the two pictures, and the difference is taken as input of SVM. Thirdly, use the training samples to train SVM. Finally, testing samples are input to classify live bird pictures and dead bird pictures. Experimental results showed that the correct identification rate exceeded 95%.