Due to the rapid development of chip technology and deep learning revolution, many ship detection frameworks for synthetic aperture radar (SAR) imagery based on convolutional neural networks (CNNs) have been proposed and achieved great success. However, there are problems hampering their development: 1) For the SAR ship detection task, it is uneconomic to apply heavy backbone network to extract features because it results in heavy computing load and prolongs the inference time cost; 2) The anchor-based methods usually have massive hyper-parameters, which typically need to be tuned carefully and easily lead to weak detection performance. To alleviate the problems, an efficient low-cost ship detection network for SAR imagery is proposed in this paper. Firstly, a simplified U-Net as the backbone to extract features is proposed. It only contains ∼ 0.47 million learnable weights, which is 2.37%, 0.76%, 0.34%, 1.01%, 0.55% and 1.07% of DarkNet-19, DarkNet-53, VGG-16, ResNet-50, ResNet-101 and ResNext-101, respectively. Secondly, an anchor-free SAR ship detection framework consisting of a bounding boxes regression sub-net and a score map regression sub-net based on simplified U-Net is proposed. To evaluate the effectiveness of our method, extensive experiments have been conducted and a more comprehensive set of evaluation metrics have been applied. Results demonstrate that the proposed network achieves 68.1% average precision and 67.6% average recall on the SAR ship detection dataset (SSDD), respectively. Compared with the state-of-the-art works, our proposed network achieves very competitive detection performance and extreme lightweight (∼ 0.93 million learnable weights in total).

This study provides new perspectives on urban development and conservation by exploring the spatial interaction between ecosystem services and urbanization. Limited studies have discussed the interaction between ecosystem services and urbanization; therefore, in this research, the spatial relationship between ecosystem services and urbanization is explored by taking the urban agglomeration in Central Yunnan as an example, and land-use data and economic and social data from 2009 and 2018 are used to determine the interactive impact of urbanization on the urban ecosystem. It is shown that (1) the spatial distribution of the urbanization level of the urban agglomeration in Central Yunnan has significant regional differences, showing a decreasing trend from the urbanized area to the surrounding areas. (2) Another factor with obvious regional differences is the spatial distribution of ecosystem services, which is similar to urbanization in spatial distribution. This difference is mainly caused by the impact of the urbanization level and the change in land use. (3) The spatial distribution and local agglomeration of urbanization and the ecosystem service value of urban agglomeration in Central Yunnan are very similar, and there is a significant negative correlation between the urbanization level and ecosystem service value. The research results have guiding significance for future urban and ecological development in Central Yunnan city.

Abstract:In order to study the K -means algorithm for evaluation of soil fertility, solve the large amount of calculation and high time complexity of the algorithm，this paper proposes the K-means algorithm based on Hadoop platform.First, K-means algorithm is used to cluster for Nongan town soil nutrient data for nine consecutive years;clustering results show that : the accuracy rate increased year by year, and consistent with the actual situation.Then for the K-means clustering algorithm in processing large amounts of data has the disadvantages of high time complexity, This paper uses the K-means algorithm Based on Hadoop platform to realize the clustering analysis of soil fertility of large amounts of data;the results show that: compared with the traditional serial K-means algorithms, improves the operation speed. The above analysis shows that, K-means algorithm is an effective soil fertility evaluation method;Based on Hadoop platform of parallel K-means algorithm has great realistic meaning to analysis of large amount of data of soil fertility factors.

In the actual classification problems, As a result of lack of clear boundary information between classification objects, that could lead to loss of classification accuracy easily. Therefore, this article from the spatial patterns of the sample properties to proceed, fuzzy clustering algorithm is proposed based on the sensitivity of attribute weights, through using the attribute weights to improve the classification capability between confusing samples, That is for researching and analysising soil nutrient spatial data with consecutive years to collect in Nongan town. Then through the analysis of the visualization technology to realize the visualization of the algorithm. Experimental results show that introducing weights portray attribute information could reduce the objective function value, and effectively alleviate the the phenomenon of boundary data that cannot distinguish. Ultimately to improve the classification accuracy. Meanwhile, use of MATLAB to form visualization of threedimensional image. The results provide a basis for to improve the accuracy of data classification and clustering analysis of large and complex agricultural data.