Applied Mechanics and Materials Vols. 530-531

Abstract: Computer tomography image (CT Image) segmentation algorithms have a number of advantages. However, most of these image segmentation algorithms suffer from long computation time because the number of pixels and the encoding parameters is large. We optimized the k-means clustering program with MATLAB language in order to improve the efficiency and stability of k-clustering algorithm in CT image segmentation. One hundred CT images are used to test the proposed method code and compare with the k-means function of the MATLAB R2012a Statistics Toolbox. We analyzed the difference of the two kinds program running time using single factor analysis of variance (ANOVA) and observed the efficiency and robustness of the segmentation results. The experimental results show that the optimized k-means clustering algorithm code has higher efficiency and robustness of segmentation. High performance of the proposed k-means clustering program is illustrated in terms of both the evaluation performance and computation time, compared with some current segmentation methods. It is empirically shown that the proposed k-means clustering program is robust and efficient for CT images segmentation.

Abstract: Image processing is the basis of computer vision. Aiming at some problems existed in the traditional image fusion algorithm, a novel algorithm based on shearlet and multi-decision is proposed. At first we discussed multi-focus image fusion and then we use Shearlet transform and multi-decision for image decomposition high-frequency coefficients. Finally, the fused image is obtained through inverse Shearlet transform. Experimental results show that comparing with traditional image fusion algorithms, the proposed approach retains image detail and more clarity.

Abstract: At present, image fusion universally exists problem that fuzzy edge, sparse texture. To solve this problem, this study proposes an image fusion method based on the combination of Lifting Wavelet and Median Filter. The method adopts different fusion rules. For the low frequency coefficient, the low frequency scale coefficients have had the convolution do the square respectively to get enhanced edge of the image fusion. Then the details information of original image is extracted by measuring region characteristics. For high frequency coefficient, the high frequency parts are denoised by the Median Filter, and then neighborhood spatial frequency and consistency verification fusion rule is adopted to the fusion of detail sub-images. Compared with Weighted Average and Regional Energy , experimental results show that edge and texture information are the most. Method in study solves the fuzzy edge and sparse texture in a certain degree,which has strong practical value in image fusion.

Abstract: To filter salt and pepper noise and protect the texture details of images effectively, an improved method of adaptive median filter is proposed. It can detect the suspicious noise by adjusting the filter window size and adopting the filter algorithm of adaptive texture direction in low density noise area and the filter algorithm of euclidean distance weighted average in high density noise area. Experimental results show that this method has better de-noising and detail-preserving performance.

Abstract: Fuzzy ontology can be built to effectively deal with uncertainty and ambiguity for domain knowledge modeling. Merging multiple fuzzy local ontologies may implement semantic integration of multiple data sources and semantic interoperability between heterogeneous systems in distributed environment. In order to solve the problem of semantic inconsistency mappings for fuzzy ontology merging system, we proposed a detection algorithm of semantic inconsistency mapping which includes sub detection methods of circular semantic inconsistency, subclass-of axiom redundancy semantic inconsistency, attribute membership semantic inconsistency and disjoint axioms redundancy semantic inconsistency. With the detection algorithm of semantic inconsistency, we establish fuzzy ontology merging system in experiment.

Abstract: The original DR image is decomposed into different scale and frequency of the band image sequence by using Laplace gaussian pyramid model methods. Using multi-scale image enhancement algorithm to enhance the High frequency component of the decomposed image, Then adjust the light of the low frequency part to make the reconstructed image illumination contrast more reasonable. The enhanced process according to different frequency layer image feature make the different gain weight for the different frequency layer image characteristics,so different frequency image layer realize respectively noise smoothing, dimensionality reduction and enhance the effect of edge character.The simulation experiments showed that this Image Processing Algorithm effect is very good.

Abstract: A major concern in aerial video surveillance today is to improve the processing speed continuously. This work introduces a joint visual constraints method for high speed aerial video stitching. To obtain real-time and accurate performance, we first select the compact and efficient ORB (Oriented FAST and Rotated BRIEF) for fast local feature description and matching, and then adopt the dynamic key frame based video stitching framework to reduce the accumulation errors. To further achieve high speed performance, we improve the above framework by fusing spatial and temporal constraints of ORB keypoints during online stitching, and increase the processing speed to 150 fps successfully. The results with public available UAV datasets demonstrate the superiority of joint constraints.

Abstract: By proposing a numerical based method on PCA-ANFIS(Adaptive Neuro-Fuzzy Inference System), this paper is focusing on solving the problem of uncertain cycle of water injection in the oilfield. As the dimension of original data is reduced by PCA, ANFIS can be applied for training and testing the new data proposed by this paper. The correctness of PCA-ANFIS models are verified by the injection statistics data collected from 116 wells inside an oilfield, the average absolute error of testing is 1.80 months. With comparison by non-PCA based models which average error is 4.33 months largely ahead of PCA-ANFIS based models, it shows that the testing accuracy has been greatly enhanced by our approach. With the conclusion of the above testing, the PCA-ANFIS method is robust in predicting the effectiveness cycle of water injection which helps oilfield developers to design the water injection scheme.

Abstract: In order to scientifically and accurately evaluate power information system, the new power information risk evaluation method based on the genetic algorithm and BP neural network is presented. The method combining the genetic algorithm and BP algorithm can be used to train the feedforward neural network , namely, first , to use the genetic algorithm to do the global training, then ,to use BP algorithm to do local precise training ,which not only overcomes the drawbacks of the traditional BP network (the training time is long, and the network is easy to fall to local extremum),but also improves the global convergence efficiency. The method was adopted to evaluate the power information system. And findings identify that the new method has distinctive convergence speed and high predicition accuracy, which provides a new concept for power information system risk assessment.

Abstract: A bio-inspired visual neural network named ICMNN was adopted to extract image structural information for image scrambling evaluation. In order to describe image structure more effectively with ICMNN, image bitmap was introduced into ICMNN input field. First, the original image was decomposed into eight binary images and each was scrambled with Arnold transformation without loss of generality. Then, ICMNN was adopted to extract the structural feature sequence of bitmap images and their corresponding scrambled ones. Last, L1 norm of the structure change sequence between them was calculated to evaluate the scrambling degree of the scrambling images. Results show that combining image bitmap decomposition with ICMNN can effectively evaluate image scrambling degree and describe the change of the structure information, which agrees with human visual perception. This evaluation algorithm is also independent of the scrambling algorithm and has a good versatility.