Computer and Modernization

SLS Algorithm for Solving Equilibrium Regular(k,2r)-CNF Formula

LI Zi-qi, XU Dao-yun

2019, 0(01): 1. doi:10.3969/j.issn.1006-2475.2019.01.001

Asbtract ( 221 )

PDF (719KB) ( 117 )

References | Related Articles | Metrics

The research on solving algorithms and structural properties of satisfiability problems is one of the important issues in computer science. In order to find some effective methods or algorithm improvement methods for some subclass problems of CNF formula, some restrictions are imposed on the structure of the formula. It is a common treatment to limit the length of a clause to a constant and the number of occurrences of arguments. The problem of satisfiability of structured formulas with regular structure and equalization of positive and negative occurrences per argument is studied. The construction of stochastic generation models and random experimental tests are helpful to observe the distribution of solutions. Moreover, the stochastic local search algorithm has good efficiency in solving the instance of a CNF formula with a certain regular structure. This paper focuses on the solution of the equilibrium regular (k, 2r)-CNF formula, that is, to solve the satisfiability problem under the condition that the length of each clause is limited to k, and the number of occurrences of each argument is exactly 2r, and the number of positive and negative occurrences of each argument is equal. The BR(n, k, 2r) model is given, this model is used to generate an equilibrium regular (k, 2r)-CNF formula with special structure, and the problem is solved by a stochastic local search algorithm. By limiting the initial assignment of 0 word and 1 word to half and uniformly generating, the comparison between WalkSAT algorithm and NSAT algorithm shows that the instance of equilibrium regular (k, 2r)-CNF formula has obvious efficiency.

Long-term Target Tracking Algorithm Based on Improved Kernel Correlation Filter

ZHANG Xue 1, NI Jian-jun 1,2, CHEN Yan1

2019, 0(01): 6. doi:10.3969/j.issn.1006-2475.2019.01.002

Asbtract ( 202 )

PDF (3952KB) ( 133 )

References | Related Articles | Metrics

Aiming at this problem that the tracking algorithm of kernel correlation filter cannot track the target accurately again after the loss of the visual target because of the occlusion, an improved kernel correlation filter tracking algorithm based on GM(1,1) grey prediction model and interval template matching is proposed. The experimental results show that the proposed algorithm has a great improvement compared with the traditional kernelized correlation filter tracking algorithm in the complex environment. At the same time, it has some advantages over other state-of-the-art methods.

Multi-scalar Multiplication Algorithm for Elliptic Curve Based on MBNS and Sliding Window

LI Yan-mei1, YIN Xin-chun1,2,SHAO Meng-li2

2019, 0(01): 11. doi:10.3969/j.issn.1006-2475.2019.01.003

Asbtract ( 323 )

PDF (857KB) ( 218 )

References | Related Articles | Metrics

Scalar multiplication heavily determines the overall implementation efficiency of Elliptic Curve Cryptography(ECC), some elliptic curve cryptosystems of public keys require multi-scalar multiplication. Multi-base number system is very suitable for efficient computation of scalar multiplications of elliptic curves because of shorter representation length and less Hamming weight. In order to improve the efficiency of ECC, this paper proposes an efficient multi-scalar multiplication based on the existing scalar multiplication algorithm in binary fields and prime fields. This new algorithm is a combination of sliding window method and multi-base scalar multiplication algorithm. The experimental results show that the new algorithm costs less compared with Shamir’s trick and interleaving with NAF’s method. The new approach can effectively improve the efficiency of scalar multiplication algorithm, so that the scalar multiplication is more efficient. Compared to other algorithms, the new approach is improved about 7.9%~20.6%.

Improved Strategy of Iterative Closest Point Algorithm

(咸阳师范学院教育科学学院,陕西咸阳712000)

2019, 0(01): 17. doi:10.3969/j.issn.1006-2475.2019.01.004

Asbtract ( 179 )

PDF (2055KB) ( 97 )

References | Related Articles | Metrics

Iterative Closest Point (ICP) algorithm is the most common method of point cloud registration. Although its registration accuracy is high, the convergence speed is slow, and the registration effect of cloud with noise and low overlapping is not good. In view of this, three improved ICP algorithms are proposed in this paper. Aiming at the point cloud with noise, the probability ICP algorithm is used to suppress the influence of the noise points to the registration results and improve the registration accuracy. In order to improve the registration speed of the point cloud, the coordinate ICP algorithm is used to realize the rapid registration of the point cloud. Aiming at the low overlapping point clouds, the box ICP algorithm is used to improve the registration accuracy and speed. The registration experiment of rabbit point cloud shows that the three improved ICP algorithms have greatly improved the accuracy and speed of point cloud registration, and all are effective point cloud registration methods.

An Improved Segmentation Algorithm for Overlapping Leaves of Plants

ZHANG Ning, WANG Zhi-ming, ZHENG Jian

2019, 0(01): 21. doi:10.3969/j.issn.1006-2475.2019.01.005

Asbtract ( 251 )

PDF (3245KB) ( 120 )

References | Related Articles | Metrics

An improved marked watershed segmentation algorithm based on HOG descriptors is proposed for overlapping leaves of tomato plug seedlings. Firstly, the leaves region and background region are segmented by using the super green transform and the OTUS to determine the overall contour of the leaves. Then the gradient of the image is calculated by using the morphological color image gradient calculation method. Using the HOG descriptor of the image gradient, the region with leaves edges and the region without leaf edges are selected, the areas without leaves edges are used as markers for watershed segmentation after morphological manipulations. Finally, image gradients are reconstructed by markers and watershed segmentation is performed to obtain segmented results of overlapping leaves. The experimental results show that the proposed algorithm can accurately select watershed markers, realize blade segmentation, and provide support for the classification of plug seedlings.

Fabric Defect Detection Based on Image Reconstruction with Auto-encoder

OU Qing-fang, XIE Huo-sheng

2019, 0(01): 27. doi:10.3969/j.issn.1006-2475.2019.01.006

Asbtract ( 232 )

PDF (2214KB) ( 120 )

References | Related Articles | Metrics

The fabric defect detection with periodic pattern generally needs to calculate period, which is not suitable for pure texture fabric. In this paper, the detection method for two kinds of fabric is proposed. Firstly, the size of detect block is set, the image block is extracted randomly from non-defect images, and then the auto-encoder is trained. Secondly, according to the size, the test images are divided into several blocks, and reconstructed with the auto-encoder, and then the mean square errors are computed between the reconstructed result and original data. Lastly, outliers of computing results of test images are detected. With rather large mean square errors value, the blocks are defect blocks. The experiments show that the proposed method is suitable for two kinds of fabric, the process is more practical and the detection results are better.

Salient Object Detection Method Based on Background Prior and Low-rank Matrix Recovery

SHEN Yang1, LI Wei1, GANG Yi-ning1, ZHAO Rui1, HAO Yue-dong2, WANG Chao3

2019, 0(01): 33. doi:10.3969/j.issn.1006-2475.2019.01.007

Asbtract ( 167 )

PDF (2651KB) ( 99 )

References | Related Articles | Metrics

Saliency object detection means that the computer automatically recognizes the saliency object in the image through the algorithm, which is widely used in many applications such as object recognition, image retrieval and image classification. Aiming at the acquisition of low-rank transformation matrix, the processing of foreground sparse matrix and the relationship between superpixel blocks in the existing significance detection model based on sparse and low-rank matrix restoration, the existing sparse and low-rank matrix restoration model is optimized to make it better applicable to the significance detection of images. Firstly, the low-rank background dictionary is obtained according to the principle of contrast and connectivity. Meanwhile, we used three scales to split multiple feature matrix of images to obtain the foreground sparse matrix of image. Secondly, the structural constraints are made for the results of the significance graph by calculating the influence factor matrix and the confidence matrix between the neighbor pixels, and sparse and low-rank matrix recovery models are used to detect the significance of the image. Finally, the propagation mechanism of K-means clustering algorithm is used to optimize the significant graph. The experimental verification on several datasets shows that the proposed method can accurately and effectively detect saliency object.

Office Supplies Target Detection Based on Faster R-CNN

FANG Jing-jing, CHENG Jin-yong

2019, 0(01): 40. doi:10.3969/j.issn.1006-2475.2019.01.008

Asbtract ( 188 )

PDF (2184KB) ( 107 )

References | Related Articles | Metrics

The technologies of RCNN network and full convolution network framework enable target detection technology to develop rapidly. RCNN networks and full convolution network frameworks are not only fast in training, but also very fast in inference speed. Besides, it has good robustness and flexibility. In the development of artificial intelligence, the key to improve the efficiency of target detection lies in good technology. We need to get a more effective and deep feature representation, which can express complex functions simply by using the multi-layer structure of deep network. The target detection method used in this paper is first to use the regional recommended network to get the proposed location and then test, because the Fast R-CNN and R-CNN target detection algorithms have been greatly improved in the running time, so the calculation area is suggested to be a computing bottle neck of the target detection. This paper joins the feature fusion technology in the algorithm, combines the features extracted from each layer of the accumulated layer, uses the regional recommended network to extract the candidate region.The regional recommendation network and the detection network share the convolution feature of the whole graph, so that the extraction time of the candidate region can be greatly shortened and the accuracy of target detection is improved.

Improved GFL Moving Target Extraction Method Based on Texture Features

DOU Xiu-chao, LI Zhi-hua, WANG Ning

2019, 0(01): 45. doi:10.3969/j.issn.1006-2475.2019.01.009

Asbtract ( 119 )

PDF (3059KB) ( 80 )

References | Related Articles | Metrics

Based on Generalized Fused Lasso (GFL) foreground model and the texture information of video, this paper proposes a moving target extraction method based on texture features. This method uses GFL foreground model to extract foreground moving object and background. Then it extracts the texture features of foreground and background in many directions by LBP algorithm, compares the similarity of two texture features, and removes the misjudged shadow regions in the foreground， which can reduce the cast shadows due to occlusion of moving targets. The paper also introduces the misjudgment rate to describe the accuracy of model. By testing real scenes that contain cast shadows, such as squares, offices, and gymnasiums, the proposed algorithm can effectively monitor the areas where shadows are cast.

Indoor Location Algorithm Based on Two-stage Position Correction

ZHANG Xin1, LIU Jian-hang1, SHANG Yong-tao2, HE Ya1, WANG Sheng-zhi1, LI Shi-bao1, CHEN Hai-hua1

2019, 0(01): 51. doi:10.3969/j.issn.1006-2475.2019.01.010

Asbtract ( 134 )

PDF (1593KB) ( 111 )

References | Related Articles | Metrics

WLAN fingerprint location technology has become a hot topic in the field of indoor location, but the traditional location algorithm are particularly susceptible to spatial environment changes which will reduce the accuracy. To solve this problem, this paper proposes the indoor location algorithm based on two-stage position correction. This paper analyzes the influence of the conductivity of indoor air on the RSSI. On the basis of the traditional algorithms’ locating results as initial location position, this algorithm uses K-Nearest Neighbor algorithm (KNN) to construct all users’ initial location position and fingerprint mapping. On this foundation, off-line and on-line stages position correction between users by multidimensional scaling (MDS) is calculated. At the last, the initial location position is optimized with two-stage position correction and the end localization result of the target is obtained. The experimental results show that the algorithm can deal with the dynamic changes of the environment effectively and correct the location results. The average error of the traditional algorithm is reduced by more than 10% after optimization of this algorithm.

Design of Steering Parameter Measuring Instrument Based on Multi-sensor Data Fusion

CHEN Hao-long, HUANG Chun-rong, LI Qing

2019, 0(01): 58. doi:10.3969/j.issn.1006-2475.2019.01.011

Asbtract ( 142 )

PDF (1094KB) ( 99 )

References | Related Articles | Metrics

According to the complex clamping mechanism and the large measuring error of the vehicle steering wheel parameter measuring instrument, a steering parameters measuring instrument method based on multi-sensor data fusion is proposed. The method compensates the angular velocity of the gyroscope through acceleration sensor and geomagnetometer, and the data sampled by the 9-axis MEMS sensor are fused by quaternion fusion algorithm, so as to correct the influence of wheel tilt angle on measurement results. The hardware platform of steering parameter measuring instrument is designed to verify the feasibility of the algorithm, the experimental results show that the proposed algorithm can improve the accuracy of angle measurement significantly.

Deep Reinforcement Learning for Step Counting Approach

PENG Chen, HAN Li-xin

2019, 0(01): 63. doi:10.3969/j.issn.1006-2475.2019.01.012

Asbtract ( 182 )

PDF (862KB) ( 157 )

References | Related Articles | Metrics

In order to deal with the problem that the user’s behavior is often uncertain in the use of step counting software, which is easy to produce various noises and the parameters in traditional algorithms cannot be continuously optimized, this paper proposes the deep reinforcement learning for step counting approach. Taking the step counting and noise discrimination as the action of the agent, the wave peak detection method is improved in the step counting and the mean crossing peak detection method is proposed. Using the recurrent neural network to save the internal state, the feedback of the user on the step effect of the pedometer is used as the reward signal to guide the parameter optimization. The experimental results show that the proposed method has high precision and strong anti-interference ability when there is noise and the mobile phone is placed in different positions, in which the noise recognition rate is 0.9151 and the step counting error rate is 0.0623.

Chinese Collective Entity Linking Method Based on Multiple Features

FENG Jun, LIU Jing-hua, KONG Sheng-qiu

2019, 0(01): 69. doi:10.3969/j.issn.1006-2475.2019.01.013

Asbtract ( 163 )

PDF (1082KB) ( 115 )

References | Related Articles | Metrics

Entity linking is the process of mapping entity mentions in a document to their entities in Knowledge Base(KB) and plays a key role in the expansion of knowledge base. Aiming at traditional entity linking methods, which mainly utilize surface features such as context similarity and ignore the semantic correlation between co-occur mentions in a text corpus, a collective entity linking method based on multiple features is proposed. Firstly, it combines synonym list and namesake list to produce a set of candidate entities. After that, it extracts varieties of the semantic features and builds a referent graph. At last, it ranks the candidate entities and choses the top1 entity as the linking target. The evaluation on data sets of NLP&CC2013 Chinese micro-blog entity linking track shows a average accuracy of 90.97%, which is better than the state-of-art result.

A Data Clustering Method Based on Competitive Swarm Optimizer

QIN Ying-bo1, CAO Bu-qing2, DENG Chun-hui1

2019, 0(01): 75. doi:10.3969/j.issn.1006-2475.2019.01.014

Asbtract ( 147 )

PDF (860KB) ( 108 )

References | Related Articles | Metrics

Data clustering plays a very important role in intelligent information processing, but the traditional K-means algorithm is sensitive to initial clustering centers. With the development of intelligent optimization algorithm, people uses intelligent optimization algorithm to cluster data and achieve a certain effect, but it is still easy to fall into the local optimization. In this paper, the competitive swarm optimizer algorithm which has achieved good results in high dimensional optimization problem is exploited for data clustering, the powerful exploration ability of competitive swarm optimizer is used to search clustering centers for data clustering. The experimental results on the five data sets of UCI show that the competitive swarm optimizer can not only get better clustering effect but also better convergence performance than genetic algorithm and particle swarm optimization algorithm.

Data Sharing Module Based on Unreliable Communication Link

LI Hui-xin, YAO Wen-ming

2019, 0(01): 80. doi:10.3969/j.issn.1006-2475.2019.01.015

Asbtract ( 177 )

PDF (2481KB) ( 93 )

References | Related Articles | Metrics

With the gradual expansion of enterprise scale and the rapid development of networks and enterprise information technology, large and medium-sized enterprises have successively established many business systems, and higher demands have been placed on the security of data synchronization between systems. For this demand, this paper designs and implements the data sharing module between unreliable communication links based on open source RabbitMQ message queue middleware. The problem of two-way data synchronization between two data centers within an enterprise is solved. Through RabbitMQs message confirmation mechanism and persistence mechanism, the data can be delivered safely without loss. The experimental results show that the function of this data sharing module is stable, which can improve the data transmission efficiency and ensure its security.

Test Case Generation Techniques Based on Constraints Presented as Boolean Expressions

SUN Yi1, YANG Xiao-hua1, LIU Jie1, YU Tong-lan1, WU Zhi-qiang2, CHEN Zhi2

2019, 0(01): 86. doi:10.3969/j.issn.1006-2475.2019.01.016

Asbtract ( 125 )

PDF (947KB) ( 101 )

References | Related Articles | Metrics

Constraints presented as Boolean expression exist widely in software specifications and programs, these constraints can be used as models of software systems and become the basis for test case generation. This paper investigates and analyzes test case generation methods based on constraints presented as Boolean expression, which are mainly divided into constraint syntax-based testing and constraint semantics-based testing. This paper summarizes various fault classes and test strategies in constraint syntax-based testing, and compares their applicability and fault detection capability. The paper also analyzes the performance of miscellaneous methods of getting and solving constraints in constraint semantics-based testing, and introduces typical tools. Finally, the future research and development are prospected.

APT Attack Prediction Model for Power Data Network Based on Ant Colony Algorithm

LIANG Jing-liang, HUANG Jun-sheng, BAI Shu-jun, WANG Peng, LI Rui

2019, 0(01): 95. doi:10.3969/j.issn.1006-2475.2019.01.017

Asbtract ( 182 )

PDF (1632KB) ( 142 )

References | Related Articles | Metrics

Advanced Persistent Threat (ATP) continuously collects business processes and target systems of attack objects in advance by the way of multi-dimension, multi-stage and multi-object, and anonymously implements data theft of network space. The power network has the natural stability demand, it covers a wide range, involves large scale and has great loss after disaster. There exist the problems of the limited security domain of network node fragmentation and the dynamic detection of the whole domain feature in current APT attack predictions. In this paper, an ATP attack prediction model for power data network based on ant colony algorithm is proposed. By designing the global trusted system model of power network, we use manifold to spread the security boundary and link the fragmented nodes to ensure global security control. The time model of APT attack is built to realize the damage analysis of the attack to the trusted system. Attack prediction model is equivalent to ant colony pheromone, which realizes automatic tracking and adaptation of APT attack. The tests and simulations show that the new model improves prediction accuracy by 12.6%.

Improved Cuckoo Algorithm for Ethylene Industry Energy Efficiency Analysis

XU Kai1,2

2019, 0(01): 101. doi:10.3969/j.issn.1006-2475.2019.01.018

Asbtract ( 173 )

PDF (1021KB) ( 119 )

References | Related Articles | Metrics

A rigorous ethylene energy consumption analysis model has been established by considering the calculation method of ethylene comprehensive energy consumption, Index Decomposition Analysis(IDA) and energy saving potential analysis method， and an improved cuckoo algorithm is proposed to identify the parameters of the model to obtain the energy-consuming energy conversion coefficient of the ethylene production. The energy efficiency of ethylene plant is evaluated by comprehensive energy consumption and energy saving based on the identified parameters. To verify the validity of the improved algorithm, the algorithm is tested on the test function. The experimental results show that the improved algorithm has better searching ability, which is better than the particle swarm optimization and cuckoo search algorithm. Finally, based on the ethylene data example, the test results show the feasibility and effectiveness of the method. It can be used as a complement to the calculation of ethylene standard energy conversion coefficient for different scales and different technologies. It provides a more practical method for analyzing the efficiency of ethylene plant.

Elimination of Pylons from Reconstructed 3D Ground Mesh of Power Transmission Corridors

ZHOU Wei-cai1, HUANG Xiang-xiang2, LIU Bing-cai1, LIN Guo-an1, PEI Hui-kun1, JIANG Wan-shou2

2019, 0(01): 108. doi:10.3969/j.issn.1006-2475.2019.01.019

Asbtract ( 173 )

PDF (5212KB) ( 112 )

References | Related Articles | Metrics

Due to the uncertainty and incompleteness of stereo matching and automatic reconstruction, it is impossible to reconstruct the power pylons mesh completely and perfectly from Unmanned Aerial Vehicle (UAV) images. The automatic reconstructed pylon mesh is not only difficult to be utilized, but also could affect the visualization effect of manually extracted power pylon models. Thus, this paper proposes an approach to eliminate these needless pylon meshes automatically based on vector pylon model. Firstly, the spatial bounding boxes of a pylon are constructed according to vector pylon type. Secondly, according to the vector pylon position and its spatial bounding boxes, the range of the needless pylon is located in the reconstructed mesh tile. Finally, based on the ray collision checking method, an algorithm of Incomplete Pylon Detection with Adaptive Distance Constraint (IPD-ADC) is designed to automatically eliminate the damaged towers. The proposed method can adaptively process the details of the tower base and eliminate unnecessary triangular surface while retaining the tower base and other ground information. The experiments show the correctness and practicality of the proposed method.

Design of Optical Fiber Communication Network Diagram

GUO Yan1, LIU Jing1,2, LI Rui-xiang1

2019, 0(01): 114. doi:10.3969/j.issn.1006-2475.2019.01.020

Asbtract ( 136 )

PDF (1248KB) ( 65 )

References | Related Articles | Metrics

In order to improve the convenience and efficiency of information management for each communication station in a power supply bureau, this paper designs and develops a vector graph editing system for power optical fiber communication network. This system not only has the common operating functions of creating, deleting, selecting, scaling, moving and Undo/Redo, but also realizes the bidirectional data configuration function of graphics object by establishing the association with relational database. In order to improve the editing efficiency of vector graphics, we propose an algorithm for dynamic generating of broken lines by analyzing the position relationship of communication station objects connected to each other in vector graphics. The results of the test show that the designed system is stable, convenient and efficient, and can meet the requirement of vector drawing of electric power optical fiber communication network.

#br# 3D Symbol Generation for Communication Situation

SUN Yu-nan, NIE Ying, REN Fei

2019, 0(01): 120. doi:10.3969/j.issn.1006-2475.2019.01.021

Asbtract ( 155 )

PDF (6060KB) ( 106 )

References | Related Articles | Metrics

The traditional military symbols are based on 2D, the symbols are drawn on the plane and then placed on the 3D spheres, which makes it difficult to express complex relationships with obvious occlusion problems and low rendering performance and also cannot express spatial relationships. In this paper, the spherical subdivision curve and Bezier curve are used to modify the coordinate transformation in the traditional military symbol, and the method of drawing primitives on the sub-line is adopted. Sub-graphic elements are used to simulate real effects such as cyber attacks and electronic interference to improve visual recognition and the three-section Bezier curve is proved and calculated by the de Casteljau algorithm, which accelerates the calculation process of the primitives. The new military symbol can solve the problems of occlusion and spatial symbol, greatly improve the rendering efficiency of terrain symbols, and make the display of connection status more practical.

Table of Content