Transfer learning between crop types for semantic segmentation of crops versus weeds in precision agriculture

Agricultural robots rely on semantic segmentation for distinguishing between crops and weeds to perform selective treatments and increase yield and crop health while reducing the amount of chemicals used. Deep‐learning approaches have recently achieved both excellent classification performance and real‐time execution. However, these techniques also rely on a large amount of training data, requiring a substantial labeling effort, both of which are scarce in precision agriculture. Additional design efforts are required to achieve commercially viable performance levels under varying environmental conditions and crop growth stages. In this paper, we explore the role of knowledge transfer between deep‐learning‐based classifiers for different crop types, with the goal of reducing the retraining time and labeling efforts required for a new crop. We examine the classification performance on three datasets with different crop types and containing a variety of weeds and compare the performance and retraining efforts required when using data labeled at pixel level with partially labeled data obtained through a less time‐consuming procedure of annotating the segmentation output. We show that transfer learning between different crop types is possible and reduces training times for up to 80%. Furthermore, we show that even when the data used for retraining are imperfectly annotated, the classification performance is within 2% of that of networks trained with laboriously annotated pixel‐precision data.     

南海北部陆坡天然气水合物成藏特征研究进展分析

南海北部陆坡区域构造地质控制着气源、流体疏导体系、富集空间及储层物性特征,因此,东沙海域、神狐海域、西沙海槽和琼东南盆地的水合物成藏条件及控制因素具有明显差异性。东沙海域深部气体可沿断层、裂缝、不整合面、砂岩疏导层和气烟囱等通道向上运移,并形成天然气水合物,具有渗漏型水合物产出特征;神狐海域水合物成藏与规模巨大的泥底辟活动相关,并与布莱克海台天然气水合物产出特征具有相似性;琼东南盆地中央坳陷带内为天然气水合物发育的重点区,底辟、泥火山或麻坑构造与天然气水合物发育密切相关。     

Case history on failure of a 67 M tall reinforced soil slope

Construction of this 67 m high RSS was completed in December 2006. After seven years in-service, a tension crack was observed at the top of the slope. In March 2015 this RSS structure catastrophically collapsed. This RSS structure collapsed in a compound failure mode; as the failure plane passed beneath, partially behind, and partially through the reinforced soil mass. The failure plane beneath the RSS was along a shale-claystone interface. The failure surface partially behind the RSS was along sandstone bedrock with water-seeping bedding planes dipping out of the rock mass. The failure surface through the upper portion of the RSS is where the geogrid reinforcement was overwhelmed by stresses originating from underlying deformation. The RSS collapse occurred after 8.3 years in-service as the shear strength along the shale-claystone interface decreased and approached the fully softened strength. The primary causative factors of this failure are: (i) an insufficient subsurface investigation program and interpretation of data for design and detailing; (ii) insufficient specifications and construction plan details for both foundation preparation and rock backcut benching; (iii) insufficient foundation preparation and rock backcut benching during construction; and (iv) adaptations to the design made during construction.     

一种外LET柔顺半铰的动力学建模与分析

对一种外LET柔顺半铰进行建模与动力学特性分析。采用斜率不连续的绝对节点坐标法（ANCF）建立了该柔顺半铰系统动力学方程；对外力作用下半铰中心线和外轮廓的运动变形进行仿真，与实验结果对比，证明了绝对节点坐标法对柔顺半铰动力学建模的有效性；进一步分析了半铰外形尺寸变化对其变形规律的影响。     

高低孔爆破处理高陡边坡的技术分析

针对前期规划不当,大宝山露天矿877 m平台高陡边坡局部存在难以采用预裂爆破技术进行处理的情况,决定采用高低孔爆破处理技术。为了解决这一难题,首先分析了高低孔的相对内涵,并对高低孔孔深计算公式进行了简单推导,然后通过公式计算得到了高低孔的孔深。在施工阶段,为了确保低孔的实施效果,两排低孔分区域使用同段别毫秒延时雷管,晚于高孔起爆。实践表明,处理后的平台平整度良好,坡面较为稳定,对高陡边坡采用高低孔爆破处理具有可行性,效果较好,可为类似边坡处理提供新思路。     

An artificial neural network approach to inhomogeneous soil slope stability predictions based on limit analysis methods

The assessment of soil slope stability is an important task in geotechnical designs. This study uses finite element upper bound (UB) and lower bound (LB) limit analysis (LA) methods to investigate inhomogeneous soil slope stability on the basis of the conventional Mohr–Coulomb parameters. The obtained stability numbers are presented in inhomogeneous soil slope stability charts. In order to minimize manual reading errors when using the chart solutions, an artificial neural network (ANN) is employed to develop a stability assessment tool for the slopes investigated in this paper. The slope stability analysis using the ANN-based tool is convenient, and the predictions it provides are highly accurate.     

Landscape dynamics of Mediterranean montane grasslands over 60 years and implications for habitats conservation. A case study in the northern Apennines (Italy)

Widespread trends of abandonment have strongly affected Mediterranean mountains after the Second World War, triggering spontaneous recolonisation of forests. A diachronic analysis of the landscape in a Natura 2000 site (Tuscany, Italy) was carried out using digital aerial photographs (1954, 2013) and a GIS-based methodology, focusing on territories above 1300 m a.s.l. The detected variations of total areas, patch shape, patch dimensions and selected metrics showed a notable shift of the vegetation towards woody types and a drastic reduction of open grasslands, some with high conservation value, accompanied by a strong increase in patch number, surface and edge. A decrease of SDI and SEI diversity indices was also observed. Considering that fragmentation is one of the main causes threatening habitats and species, our results point out a clear necessity for a monitoring programme and suitable actions aimed at improving the status of biodiversity-rich montane grasslands.     

浅析高速公路高边坡病害原因及治理对策

对高速路建设中高边坡病害的成因展开分析,并对高边坡病害的治理提出对策,希望对高边坡病害治理有一定的促进作用。     

四川省植被覆盖时空演变及未来变化趋势分析

四川省地处长江上游,是长江流域重要的生态屏障;解析该区域植被覆盖的时空演变特征,对长江上游的生态环境修护与治理具有重要意义。基于长时间序列的MODIS NDVI数据,解析2000~2016年间四川省植被的空间分布格局与时间变化特征;采用趋势分析法和Hurst指数,预测未来植被的变化趋势及持续性。结果表明:(1)四川省NDVI多年平均值为0.50,植被覆盖整体上呈现川东高于川西的空间分布特征;(2)2000~2016年间,四川省的NDVI值在0.48至0.53之间波动,整体呈上升趋势,增长率为0.002 5/a;(3)植被覆盖度在未来变化将呈现改善大于退化的趋势;植被持续性改善的面积大于持续性退化的面积。