Conditional random field with the multi-granular contextual information for pixel labeling

To make full use of the contextual information object recognition and scene understanding, a multi-granular context conditional random field (MGCCRF) model is presented to combine context information in a variety of scales. It is efficiently implemented through extending the pairwise clique to the multi-granular context windows. In the fine-granular context window, the label consistency of similar features can be obtained with the probability of the label transferring between two adjacent pixels. At the same time, the spatial relationships among different classes in the coarse-granular context window are explicated in details. To train the MGCCRF model, a piecewise training method with the bound optimization algorithm is designed to improve the performance. Experiments on two real-world image databases show that compared with other methods, the modified conditional random field model is more competitive and effective in terms of the quantitative and qualitative labeling performance.     

An auto-encoder-based summarization algorithm for unstructured videos

Multimedia Tools and Applications - Video summarization is an effective way to quick view videos and relieve the pressure of videos storage. However the traditional algorithms are hardly adapted to...     

Multi-scale segmentation of very high resolution remote sensing image based on gravitational field and optimized region merging

Multimedia Tools and Applications - This paper proposes a multi-scale segmentation approach for high resolution remote sensing image (HRRSI) based on the gravitational field and region merging. In...     

Efficient aerodynamic shape optimization of transonic wings using a parallel infilling strategy and surrogate models

Surrogate models are used to dramatically improve the design efficiency of numerical aerodynamic shape optimization, where high-fidelity, expensive computational fluid dynamics (CFD) is often employed. Traditionally, in adaptation, only one single sample point is chosen to update the surrogate model during each updating cycle, after the initial surrogate model is built. To enable the selection of multiple new samples at each updating cycle, a few parallel infilling strategies have been developed in recent years, in order to reduce the optimization wall clock time. In this article, an alternative parallel infilling strategy for surrogate-based constrained optimization is presented and demonstrated by the aerodynamic shape optimization of transonic wings. Different from existing methods in which multiple sample points are chosen by a single infill criterion, this article uses a combination of multiple infill criteria, with each criterion choosing a different sample point. Constrained drag minimizations of the ONERA-M6 and DLR-F4 wings are exercised to demonstrate the proposed method, including low-dimensional (6 design variables) and higher-dimensional problems (up to 48 design variables). The results show that, for surrogate-based optimization of transonic wings, the proposed method is more effective than the existing parallel infilling strategies, when the number of initial sample points are in the range from N _v to 8N _v (N _v here denotes the number of design variables). Each case is repeated 50 times to eliminate the effect of randomness in our results.     

一种非侵入的GUI自动化测试系统设计

针对嵌入式GUI软件自动化测试,提出一种非侵入的GUI自动化测试系统设计方案,能够使用Python格式的测试脚本灵活描述测试用例的执行步骤,能够自动化执行测试脚本;系统将人机交互步骤转成鼠标键盘模拟设备的控制命令,实现对被测软件鼠标键盘的控制,从而输入需要的人机交互信息,实现测试用例的自动执行;同时以不干扰被测系统为原则,直接采集被测系统输出到显示器中的视频数据,通过对图像数据的处理进行图像对比、图像查找和文字识别,根据测试用例预期输出自动判断用例的执行结果;该测试系统解决了嵌入式软件GUI自动化测试的难题,提高了测试效率。     

计算机视觉技术的车型识别应用研究

针对目前立体停车库主要采用传统的传感器入库检测系统来识别车辆信息,存在施工周期长、器件损坏率较高,维护成本较高的问题,提出了一种应用计算机视觉技术的车型识别解决方案.应用卷积神经网络框架Caffe,基于Caf-feNet模型,通过fine-tuning模型优化以及参数优化,最终得到了一个性能较优异的识别模型.实验结果表明,该模型可克服输入图片背景复杂多变,目标被遮挡的情况,对轿车车型识别这一问题鲁棒性好,具有一定的可行性及应用价值.     

面向高质量加工的NURBS曲线插补算法

通过分析数控系统加工时常用的插补算法的特点,提出一种基于NURBS曲线的插补算法.该算法包括速度规划和实时插补两部分:速度规划部分考虑了工件加工时允许的最大轮廓误差,以保证高速运行过程中加速度的连续,使机床运行平稳,避免产生大的冲击;实时插补部分应用弦截法计算插补参数,能将实时插补产生的速度波动控制到理想水平,进一步减小了机床震颤.仿真实验结果表明,文中算法能够减小机床振动,实现高质量加工.     

Remote Sensing Prospecting in High Vegetation Coverage Area——A Case Study of the in Datuanbao Copper Ore Deposit

This article described the way to extract information of copper and lead anomaly in high vegetation coverage area through a case study from the Datuanbao copper ore deposit and its environs,Jiangcheng,Yunnan Province.Principal component analysis,false color composite,stretch Histogram and supervise classification etc were analyzed in the remote sensing images using ENVI4.6 software.It is suggested that the linear structure,which is delineated by convolution filter and morphological analysis method,is characterized by statistical self-similarity and fractal geometry.It is found that the high value of fractal dimension is roughly coincident with the copper and lead anomaly of vegetation by the comparison between contour maps of plant geochemical anomaly and fractal dimensions,which is calculated by box-counting.As a matter of fact,80 percent of copper and lead deposits or occurrences are located in areas of high value of fractal dimension and vegetation anomaly.By the comprehensive analysis of stratigraphy,igneous rocks,copper and lead mineralizations,fractal dimension of linear structures and vegetation anomaly from remote sensing,nine exploration targets with three levels of probabilities are figured out.     

Stochastic feedback and its relation to the entropy criterion

In this paper, we generalize the interpretation of the entropy criterion by using stochastic feedback. When a common white Gaussian noise of zero mean and unit variance is applied, the entropy of a transfer function is shown to be equivalent to the maximum variance of the outputs fed back and added to the input through multiplication of a stochastic gain. It turns out that the existing minimum entropy control is useful for minimizing the worst effect of external noises on the outputs when stochastic feedback is applied.     

A method to exchange procedurally represented 2D CAD model data using ISO 10303 STEP

Procedural models have the advantage of being easy to edit simply by changing the values of the parameters of their constructional operations. Such models are said to embody design intent, in the sense that modifications to them conform to the method of creation used by their original creator. They also comply with any constraints implied by the particular constructional operations used. This paper introduces the development and standardization process of the ISO 10303-112 specification and describes the concept of procedural 2D modeling, a method of representing procedural 2D CAD models in STEP in harmony with other STEP resources. The feasibility of procedural 2D modeling commands for the exchange of procedurally represented 2D CAD model data is demonstrated through an experiment where procedural 2D CAD model data in neutral form generated with an in-house 2D modeling system are translated to and modified in a commercial 3D mechanical CAD system.