It’s Time To Learn About Learning: Where Should the Environmental and Natural Resource Governance Field Go Next?

This Commentary reflects on the state of the scholarship on learning for environmental and natural resource policy and governance. How have we been learning about learning? We highlight theoretical and empirical advancements related to learning, as well as areas of divergence between learning theories and frameworks, and underdeveloped knowledge around processes and outcomes. To address these limitations and improve progress in both theory and practice, we offer recommendations for learning scholarship by focusing on how to collectively engage in ‘learning about learning’.     

警用视频监控的犯罪防控效果及空间差异 ——以苏州市姑苏区为例

采用双重差分法,对苏州市姑苏区2014~2016年的警情数据进行分析,评估警用治安视频监控的犯罪防控效果。结果表明,视频监控对犯罪具有明显的抑制性作用。在案件类型上,视频监控对盗窃类案件的防控效果较好,但对盗窃电动自行车、盗窃电动车电瓶的防控效果较小。在时间维度上,与节假日相比,对工作日犯罪的抑制效果更好;与晚上相比,对白天犯罪的抑制效果更好。针对通过局部空间自相关分析所获取的重点区域,与警务人员访谈和实地调研后发现,视频监控的犯罪防控效果与周边地理环境、人流密集程度、警务情况等密切相关。     

Farmer-based Rice Cropping Systems in the Poyang Lake Region,China: Analysis of Characteristics and Policy Implications

This paper examines the spatial characteristics of farmer/household behaviors in regional rice cropping systems (RCS), and the results provide necessary information for developing strategies that will maintain regional food security. Through field study and statistical analysis based on 402 households questionnaires finished in 2014-2015 in the Poyang Lake Region (PLR) of China, we arrived at two main conclusions. First, single- and double cropping rice were found across the study area, but showed a general distribution trend, with double cropping rice in the southeast part (especially in Jinxian county) and single cropping rice in the northwest (particularly in De’an county). Second, the household decisions concerning RCS varied in different parts of the PLR, but double cropping was the dominant type, with about 63.57% of the respondent households in the PLR cultivating double cropping rice. However, the multiple-cropping index of paddy rice was only 1.55. About 3% of interviewed households had altered their RCS during this period. Based on these findings, the local governments should guide farmers’ paddy field cultivation behaviors by increasing the comparative efficiency of rice production, promoting appropriate scale operations and land conversion, as well as optimizing rice growing conditions to improve the multiple cropping index and enhance food provision. Finally, land-use efficiency and more sustainable use of land resources should be improved.     

基于EEMD和ARIMA的海温预测模型研究

类型丰富、时空分辨率高的海洋探测数据,为信号分解和机器学习算法的应用提供了可能。本文针对如何建立有效的海温预测模型这一问题,使用高时空分辨率的海表温度(SST)融合产品,引入信号处理领域的集合经验模态分解(EEMD)和机器学习领域的自回归积分滑动平均模型(ARIMA)。首先利用最适于分解自然信号的EEMD方法,将海温数据分解成多个确定频率的序列;再利用ARIMA分别对各个频率的序列进行预测,最后将各个序列的预测结果进行组合。该方法在丰富数据的支撑下,比以往直接使用海温数据所建立的预测模型精度更高,为更好地进行海温预测提供了新方法。     

Development strategies for the sea cucumber industry in China

The sea cucumber, Apostichopus japonicus, is an important marine aquaculture species in China. After nearly thirty years of development, the production of A. japonicus has become commercially lucrative and successful. In this report, current advances in sea cucumber industry are addressed in terms of the basic biology, culturing methods, and health care bene?ts. Next, the challenges restricting development of the sea cucumber industry are discussed, including weaknesses in the basic biological research, the problem of germplasm degradation, environmental stress caused by global climate change, and food safety problems. Finally, several strategies are presented that might contribute to sustainable development of the sea cucumber industry. These strategies include advances in genome studies, behavioral studies, selective breeding, ecological culture technologies, reforms in food safety management, and the development of health care functions based on contemporary medical practices. Thus, our work provides new insights into how to explore the sustainable development of the sea cucumber industry in the future.     

加权空-谱联合保持嵌入的高光谱遥感影像降维方法

高光谱遥感影像数据量大、波段数多,容易导致“维数灾难”。传统流形学习方法一般仅考虑其光谱特征,忽略了空间信息。为此提出一种非监督的基于加权空-谱联合保持嵌入(WSCPE)的维数约简算法。首先采用加权均值滤波(WMF)方法对高光谱影像进行滤波,以消除噪点和背景点的干扰。然后根据遥感影像地物分布的空间一致性,通过采用加权空-谱联合距离(WSCD)来融合像素点的光谱信息和空间信息,有效选取各像素点的空-谱近邻,并根据像素点与其空-谱近邻点之间的坐标距离来有区别的利用其近邻点进行流形重构,提取低维鉴别特征进行地物分类。在PaviaU和Indian Pines数据集上的分类结果表明,总体分类精度分别达到了98.89%和95.47%。该方法在反映影像内部流形结构的同时,有效融合了影像的空间-光谱信息,故能提高影像特征的鉴别性,并提升分类性能。     

高分辨率光学遥感场景分类的深度度量学习方法

针对高分辨率光学遥感影像场景具有同类型内部差异大、不同类型间相似度高导致部分场景识别困难的问题,本文提出了一种深度度量学习方法。首先在深度学习模型的特征输出层上为每类预设聚类中心,其次基于欧氏距离方法设计均值中心度量损失项,最后联合交叉熵损失项以及权重与偏置正则项构成模型的损失函数。该方法的目标是在特征空间上使同类型特征聚集并扩大类型间的距离以提高分类准确率。试验结果表明,本文方法有效地提升了分类准确率。在RSSCN7、UC Merced和NWPU-RESISC45数据集上,与现有方法相比,分类准确率分别提高了1.46%、1.09%和2.51%。     

结合深度学习和图割法的遥感影像建筑物检测

提出一种基于卷积神经网络和图割法的自动提取高分影像建筑物的方法。首先,通过卷积神经网络定位与检测建筑物的位置,逐一提取单个建筑物轮廓,利用检测结果分别建立建筑物和非建筑物的高斯混合模型（GMM）,然后结合最大流最小割的图像分割方式实现全局优化,完成建筑物初步提取,最后用形态学进行优化。通过试验证明了该方法的可行性。     

结合主动学习和词袋模型的高分二号遥感影像自动化分类

高分卫星遥感影像空间分辨率的提高,使得地物的光谱和纹理变得更加丰富和复杂,这给遥感影像的自动化分类带来严重挑战。因此,本文提出了一种结合主动学习和词袋模型的高分二号遥感影像分类方法。首先,对研究区域进行多尺度分割,建立影像分割对象集;然后,采用词袋模型构建影像对象的语义特征向量;最后,充分考虑位于分类边界的不确定性样本分布,迭代选择最优样本用于训练支持向量机,用于分类遥感影像。为了验证本文方法的有效性和稳健性,以山东省某市的高分二号遥感影像为试验数据进行了试验分析。结果表明,本文提出的方法可以有效地将研究区域分为水体、地面、植被和建筑物四类,正确率达到90.6%以上。     

新工科理念下的实验课程教学改革与实践探索

近年来为适应创新人才的培养,"卓越计划"开启了全新教学模式。基于新工科理念,为了更好地进行实验课程教学,以"3ds Max软件三维建模"为例,笔者从课程内容安排、教学方式、考核机制等方面进行了教学探索和实践。本文以应用软件实验类课程项目教学的实践经验为基础,注重学生团队合作和交流能力、实践能力、创新能力培养,以及多渠道快速学习新事物的能力,通过与测绘新技术下多软件融合,使学生拥有知识整合能力,为提高实践教学提供参考和借鉴,以更好地适应新工科理念下的创新人才培养。