A learning-based method to detect and segment text from scene images

This paper proposes a learning-based method for text detection and text segmentation in natural scene images. First, the input image is decomposed into multiple connected-components (CCs) by Niblack clustering algorithm. Then all the CCs including text CCs and non-text CCs are verified on their text features by a 2-stage classification module, where most non-text CCs are discarded by an attentional cascade classifier and remaining CCs are further verified by an SVM. All the accepted CCs are output to result in text only binary image. Experiments with many images in different scenes showed satisfactory performance of our proposed method.     

对小波图像去噪方法的探讨

小波图像去噪方法已广泛应用于图像去噪领域．在对日前小波图像去噪文献进行综合的基础上，描述了信号奇异性，并阐述了三种基于小波阈值去噪的方法，最后对小波去噪方法进行了一些展望．     

一种在MPEG-4编码过程中提高实时性的算法

本文提出了一种新的在MPEG-4编码过程中提高实时性的算法。算法采用双重筛选机制,充分利用了运动估计的结果提取运动对象。首先在运动估计过程中加入背景优先准则进行第一重筛选;然后根据运动估计后宏块的最小平均绝对差值分利用区域自适应门限进行第二重筛选,从而获得运动区域掩模;最后对掩模进行形态滤波和填充消除噪声点,最终得到运动区域。算法分别在硬件和软件平台上进行测试,结果表明提出的算法得到了理想的分割效果,较传统的在编码前进行运动对象提取的方法实时性大大提高了。     

A framework of region-based dynamic image fusion

INTRODUCTION Image fusion is a sub-area of more general topics of Information Fusion and Data Fusion and deals with static images and dynamic images (image sequences or video data) (Maitre and Bloch, 1997). By com-bining the information from a range of image data obtained by different sensors or the same sensor with different imaging schemes, the composite image al-ways presents the complementary information of different input images at the same time and provides us the integrated descr…     

Video diaries: a tool to investigate sustainability-related learning in threshold spaces

This paper takes as a case study an undergraduate field class from a UK university to rural Uganda. It describes and evaluates the use of video diaries as a tool for investigating the process of transformative learning in the context of education for sustainability. The applicability of threshold concept theory to this learning is investigated. Results show the video diary technique to be much more successful in capturing the development of student learning than written diaries. Despite some practical difficulties, rich data were generated. Analysis demonstrated that learning met three of the five criteria for threshold concepts: the learning was transformative, integrative and troublesome. Further research could test whether the learning in this case study met the threshold concept criterion of irreversibility. Further elaboration of the theoretical relationship between threshold concepts, transformative learning and education for sustainability is also needed. It is concluded that the video diary method is potentially transferable to other educational research contexts and may be particularly suitable for researching learning in threshold spaces.     

基于CFAR的SAR图像分割

选用瑞利分布模型对SAR图像进行恒虚警率(CFAR)检测。提出新的参考窗口选取方法,并针对虚警特点提出新的均值估计方法。实验证明,与一般的均值估计方法相比,文中提出的均值估计方法是可行的、有效的,且与其中提出的参考窗口选取方法相结合能更好地去除虚警,保留目标信息。     

城市人口比重与非农产品消费比重关系研究

城市化是一个复杂的过程,城市的经济发展是实现这一过程的重要推动力之一.城市是非农产品的生产地,因此城市人口比重与非农产品支出比重的提高对于城市经济的发展会产生直接的带动作用.明确二者之间的关系,对于我国稳步推进城市化进程具有重要意义.     

高校扩招对我国居民收入分配差距的影响研究——基于1999—2018年省际面板数据的实证分析

本研究围绕高校扩招对居民收入分配差距的影响进行实证研究,选取1999—2018年我国31个省份(不含我国港、澳、台地区)的面板数据进行门槛回归估计。本研究发现,样本期间内高校扩招对于缩小收入分配差距具有显著作用,且表现出鲜明的门槛特征。高校扩招对收入分配差距的影响还存在经济门槛效应和技术门槛效应:经济发展水平越高,高校扩招缩小收入分配差距的作用越强;随着技术水平的提升,高校扩招缩小收入分配差距的作用先增强后减弱。据此,为更好地推动高等教育与社会经济发展,发挥高等教育人力资本促进居民收入分配公平的作用,我国可以实施因地制宜的分地区高校扩招政策;强化经济发达地区对经济欠发达地区的带动作用;加快培养高层次应用型人才以适应技术进步对高层次人才的需求。     

环云南地区入境旅游与经济发展的空间溢出及门槛效应

随着云南成为我国入境旅游中心,本文基于2006-2017年云、贵、川、渝、藏、桂六个地区面板数据,综合运用空间计量回归和门槛面板回归模型深入分析了六个地区入境旅游的空间相关性、空间外溢和门槛效应入境旅游对经济发展存在着非线性的空间入境旅游收入增长、入境旅游人数增长对经济增长有着显著的门槛效应。随着增长率提高,其对经济发展的弹性在减少。因此,对于经济发展水平相对低的云南、贵州,可以通过发展入境旅游来促进其经济快速增长,但对于四川、贵州、西藏,由于门槛效应的存在,需要优化旅游产品和旅游空间结构,以促进入境旅游对经济发展的作用。