Benchmark datasets are essential to develop and evaluate remote sensing image retrieval (RSIR) approaches. However, there are no two datasets with different remote sensing image sources, an approximate equal number of images, the same classification system, and image size in the existing public benchmark datasets. This may affect the evaluation of the universality and robustness of RSIR approaches for the same category in different datasets, and even hinder the development of new cross-source RSIR approaches that require remote sensing images from different sources. We therefore present two new large-scale datasets from ArcGIS and Bing World Imagery, respectively. Similar to the PatternNet dataset from Google Earth Imagery, both two new collected datasets contain 38 classes using the same classification system and each class has at least 1,500 images. We conduct experiments using five handcrafted low/mid-level feature methods and six deep learning high-level feature methods on the two datasets. Results show that our datasets are effective for evaluating different RSIR approaches and the results can be served as the baseline for future research. We also perform the comparison and the cross analysis compared with other large-scale datasets. Results indicate that our datasets are more inclusive, richer variations and better intra-class diversity. Besides, other experimental results show that our new datasets and the VGoogle (extracted by volunteers from Google imagery) dataset can be merged into one dataset for larger-scale remote sensing image retrieval.

In order to improve the dynamic optimization of fleet size and standardized management of dockless bike-sharing, this paper focuses on using the Markov stochastic process and linear programming method to solve the problem of bike-sharing fleet size and rebalancing. Based on the analysis of characters of bike-sharing, which are irreducible, aperiodic and positive-recurrence, we prove that the probability limits the state (steady-state) of bike-sharing Markov chain only exists and is independent of the initial probability distribution. Then a new “Markov chain dockless bike-sharing fleet size solution” algorithm is proposed. The process includes three parts. Firstly, the irreducibility of the bike-sharing transition probability matrix is analyzed. Secondly, the rank-one updating method is used to construct the transition probability random prime matrix. Finally, an iterative method for solving the steady-state probability vector is therefore given and the convergence speed of the method is analyzed. Furthermore, we discuss the dynamic solution of the bike-sharing steady-state fleet size according to the time period, so as improving the practicality of the algorithm. To verify the efficiency of this algorithm, we adopt the linear programming method for bicycle rebalancing analysis. Experiment results show that the algorithm could be used to solve the disordered deployment of dockless bike-sharing.

In recent years, with the continuous development of online geographic information services, the multiscale online expression of geographic elements and topological information has become a hot research topic. Currently, a large amount of research focuses mainly on the compression of vector line elements. The main shortcoming of current algorithms is that vector data compression does not consider the optimality of line simplification. In the face of terabytes or even petabytes of data, problems arise such as low processing efficiency and long transmission time. Based on the shape-tree principle, the proposed method aims to reconstruct the organizational structure of a geographic element and establish a hierarchical representation model of the element based on its structure, realizing multiscale shape recognition and matching and searching geographic elements. Finally, the proposed method is compared with a model based on the wavelet transform and discrete Fourier transform. The results demonstrate the applicability and effectiveness of the proposed method in morphological expression with geospatial big data.

Since the United Nations passed the 17 Sustainable Development Goals (SDGs) and 169 sub-goals in 2015, many researchers have tried their best to put them into practise. However, mass data and different understanding of SDGs indicators make this work much more difficult. In this article we reference the concept of essential variables which may have some contribution to established a standard system for SDGs monitoring. Here we take SDG15 as an example. Firstly, we deeply analyse the main contents of SDG15 and select some geospatial-related indicators involved with forest land. From those forest-related and geospatial-enabled indicators we concise 5 essential forest variables (EFV4SDG15) which could be classified into 3 different types: land cover distribution and transformation, aggregative pattern and intensity and spatial-temporal evolution process. Secondly, formalized expression of EFV4SDG15 and extraction of them from multiple data source are necessary, including data processing and mathematical expression of the variables. Finally, we take global region as research target to carry out preliminary experiments. In this part, we only select 2 EFV4SDG15s: the forest coverage rate and forest land transfer matrix to show detailed operation process and then present the results in forms of figure and table. The paper has preliminary attempts to the application of essential variables in SDGs monitoring and provides an example for other fields.