The indoor navigation network is the basis for pedestrian navigation, information recommendation, and business analysis. The traditional method of manual mapping or semiautomatic extraction of three-dimensional indoor navigation network cannot meet the requirement of high-frequency change of complex indoor space structures. With the continuous development of indoor positioning technology, there is an explosion of trajectory data of indoor moving objects, which provides a possibility for rapid construction and change monitoring of indoor navigation networks. This paper proposes a method of crowdsourcing construction of indoor navigation network based on the trajectory of moving objects. Based on trajectory simplification preprocessing using ST-DBSCAN, an indoor trajectory adaptive rasterization algorithm is proposed to reduce the influence of raster image resolution on the extraction of navigation networks. This approach effectively avoids the failure of navigation networks' topological connection that is caused by the difference of track trajectory density. Moreover, it automatically identifies the connection points between floors by the CFSFDP adaptive clustering algorithm to realize the rapid construction of indoor navigation networks. The experimental data is derived from the real indoor moving object trajectory data provided by Shanghai Palmap Science & Technology Co., Ltd. The experimental results show that, compared with the universal rasterization method, the proposed method improves the accuracy of navigation network construction by an average of 2.43% and improves the accuracy of topology by 12.8%.
Emergency facilities play an important role in disaster prevention and mitigation. However, most of the relevant studies and practices usually focus on urban areas. Villages and towns in China are vast in territory and complex in environment. Because of the low level of economic development in the past and out-dated emergency facilities developed by traditional layout methods in China, these places become vulnerable to disaster . In this case, quantitative study on the layout evaluation and optimization of emergency facilities in villages is necessary. With the view of “the demand points distribution - identifying facilities alternative sites based on the suitability evaluation for emergency facilities - model construction and application-rationality evaluation of site selection schemes and planning proposals”, this paper concentrates on the construction of emergency facilities and attempts to establish a relatively reasonable method for emergency facilities layout in villages. The main contents of the article are as follows: (1) establishing suitability evaluation index system for emergency facilities from geographical conditions, traffic conditions, population distribution, and hazard sources distribution. In particular, safety issues are considered to avoid the constructions in areas with high incidence of geological disasters, floods, and other disasters; (2) building location model which consists of set cover model and maximum coverage model to analyze emergency facilities that only need to undertake emergency service and building location model which consists of set cover model and P-median model to analyze emergency facilities that need to undertake both emergency service and public service; and (3) taking response coverage and cross-response rate as the efficiency evaluation indicators, while the maximum response time and average response time from demand point to their nearest facilities are used as the fairness evaluation indicators. Both indicators are used to evaluate multi-schemes generated by the location models. In this paper, Songbai Town of Shennongjia in Hubei province is taken as the research area, fire stations and emergency medical centers are used as two typical types of emergency facilities for this research. The results show that: (1) the suitability evaluation index system can comprehensively reflect various factors synthetically; and (2) the model calculation results are superior to the original schemes in terms of efficiency and fairness indicator. However, the more optimized layout schemes need to be determined together with population distribution, local situation, and cost factors. Thus, the proposed method has good feasibility and reliability.
Building reconstruction based on 3D point cloud data has broad application prospects in fields such as high precision urban mapping and virtual reality. Due to the diverse geometry of buildings, there are widespread problems in traditional reconstruction algorithms, e.g., slow computation speed, low fitting precision, and incompleteness of building structures. Thus, with single-building as the research object, this paper proposed an algorithm based on weighted constraints for reconstructing point cloud surfaces. By fully considering each point’s contribution to the fitting plane during the surface initialization process, the proposed algorithm, which is based on regular sets, simultaneously optimizes the error of adaptively weighted fitting and the smoothness of neighbor structures. The algorithm was applied to the 3D point clouds of various buildings. Results showed that, compared with conventional building reconstruction strategies, the weighted-constraints based algorithm of this study can design adaptive weights according to different types of point clouds, and can choose the optimal weight for model fitting. In cases where the point cloud data contain high noise and low accuracy, the proposed algorithm can help generate more accurate surface models for single-building.
Landscape visual resource assessment is important to effectively and reasonably evaluate the value of landscape visual sources.Previous studies on landscape visual resource assessment mainly focus on subjective evaluation and scores by experts or tourists, while few studies areon quantitative evaluation of landscape visual resources. To help close the gap, this study proposed a comprehensive evaluation method of landscape visual resources based on GIS and FAHP ( Fuzzy Analytic Hierarchy Process) to objectively reveal the distribution characteristics and influencing factors of landscape visual resources from holistic and single-factor perspectives. Specifically, the study integrated basic visual elements and landscape visual sensitivity elements (e.g., sight, viewshed, relative distance sensitivity, relative slope sensitivity, environment contrast) to design an evaluation indicator system from the perspective of comprehensive quantitative evaluation of landscape vision. Then, adopting GIS spatial econometrics analysis and FAHP, the study calculated the comprehensive visual index to objectively reveal the distribution characteristics of landscape visual resources and their influencing factors from the holistic perspective and the single factor perspective. Taking Mount Tai as the research area, the test results show that: (1)from the single factor perspective, Yuhuangding, Tianzhufeng, and Bixiaci have the highest scores in terms of viewshed, relative slope sensitivity, and relative distance sensitivity; and(2) from the holistic perspective, Yu Huang Ding has the highest visual index (0.819) and Tao Hua Yu has the lowest one (0.180). In addition, the value of visual index is also affected by altitude, distribution of spots, and routes, etc. Compared to some traditional evaluation models that relies on tourists or experts' subjective opinion, this study proposed a new perspective of objective quantitative assessment to evaluate the landscape visual resources by combining a comprehensive evaluation index system and a evaluation model. Verification results indicate that this study builds a new research idea and a research method for quantitative description and assessment of landscape visual resources, showing better reliability and advancement. Our findings also provide objective references and decision supports for the comprehensive development planning of scenic spot development and protection. Meanwhile, in this study, landscape visual resources were mainly discussed from the perspective of GIS, while the real assessment effects of landscape vision were often affected by many complex factors. Therefore, how to establish a more synthetical and quantitativeassessment system to measure and evaluate landscape visual resources is one of the most important next stepsto be done in the future.
Land desertification is one of the most serious eco-environmental problems in the world, and is among the top ten environmental and development problems that threaten the survival of human beings. China has severe desertification, and desertification has greatly affected China's eco-environment and social development. In this context, the research on desertification is of great significance to China. The evolution of desertification is mainly manifested in the change of desertification area, the area change, change of desertification types, vegetation change, and so on. In this paper, the 31 provincial regions of China (excluding Taiwan, Hong Kong, and Macau) were included as the study area. Remote sensing was utilized to study the spatiotemporal evolution of the thematic factors of desertification, salinization, and erosion-induced desertification. Following the principles of regional differentiation, genesis, and multi-level sequences, China was divided into 8 desertification zones, 42 desertification sub-regions, and 36 desertification communities. Based on the desertified land type maps of 1975, 2000 and 2017, we analyzed the desertification evolution by focusing on the changes in the area and gravity center of China’s desertification. According to the percentage of increase or decrease of desertified land area over a certain period of time, the evolution types of desertification can be divided into 7 categories. The present study reveals the current situation of China’s desertification, and analyzes the spatiotemporal evolution and the gravity center movements of different desertification regions. Next, the key target areas and management suggestions of desertification control in China were discussed. Our findings are as follows. (1) There are many kinds of deserts and desertified lands in China, of which the area of severe desertification is 25.18×104 km2, accounting for 19.59% of the total desertification area. (2) From 1975 to 2017, China's desertification has been significantly reversed, and the dynamic characteristics of desertification in 1975-2000 differed from 2000-2017. The intensity of desertification in 2000-2017 significantly reduced as compared to 1975-2000, thanks to the decrease in aggravated areas and increase in weakened areas. (3) The migration value of desertification center of gravity can indirectly reflect the development trend and degree of dynamic change of desertification, the greater the migration, the more significant the difference of desertification change in this region, and the direction of center of gravity migration is the area where desertification area increases. The coordinate migration of desertification barycenter points in different periods has the same orientation.
The presence of clouds hinders remote sensing of earth surface and causes information loss. Understanding the spatiotemporal patterns of different cloud states over a given region can improve the pertinence and accuracy of cloud noise elimination. In this study, the MODIS surface reflectance products, MOD09A1, were used to extract the cloud state information for each pixel of Hunan Province, spanning from 2001 to 2017. Then, geo-statistics and spatial analysis methods were used to detect Hunan's spatiotemporal patterns of the clear, cloudy, and mixed states. The examined patterns included the spatial distribution, seasonal distribution, gap length duration, and cloud interference for the different cloud states. The results show that: (1) Hunan Province in general was quite severely affected by clouds, with obvious spatial heterogeneity, i.e., the cloudy state when imagery are shadowed by clouds mainly occurred over the western and southern mountain areas of Hunan, while the mixed state when imagery are distorted by clouds existed over the gentle plains and hilly regions of the northern, central, and southern Hunan; (2) the cloudy state was dominant among the cloudy and mixed states, with the former more likely to occur between January and February, November and December, late May and early July, while the latter maintained 10% of cloud cover during the whole year and up to 18% from June to October; (3) cloud gap lengths of 8-days and 16-days were the primary situation for the cloudy and mixed states;(4) increasing the composite window size, cloud interference effects declined dramatically and could be neglected if monthly composite size (4 collections) was utilized;(5) by Principal Component Analysis (PCA) of the different cloudy state data, the first two PCA components were derived which indicated different spatial distribution patterns and divided Hunan into four sub regions with distinctive cloud characteristics; and (6) there are significant relationships between the different cloud states and elevation variation, and except for the cloudy state, they are all negatively correlated with elevation variation. This study can provide technical support for the selection of remote sensing data, the removal method of cloud noises, and the reconstruction of vegetation indices time series for Hunan Province.
Gridded population data can be used to describe the actual spatial distribution of populations and is an effective way to achieve better integration of population data with natural, social, and economic factors. This study analyzed the demand of fine-scale gridded population data. Taking the densely populated Shandong Province in eastern China as an example, the spatialization method of township-level demographic data was investigated using nighttime satellite data and land use data fusion modeling. In this process, EVI was used to reduce saturation of DMSP/OLS nighttime satellite data to increase the difference of population distribution within the urban land. The urban and rural two-level partition method was used to avoid the shortcomings of nighttime data in the low radiance area of rural areas. The demographic values from the rest region were used to evaluate the modeling accuracy, and the results showed that 78% of the administrative units had an absolute relative error of less than 20%. Finally, based on the population data of the township-level which was first published in the fifth census in 2000, the gridded population data SDpop2000 at 100 m- resolution in Shandong Province was generated. The SDpop2000 was compared with the global WorldPop population data product with higher precision. The results showed that the correlation coefficient between SDpop2000 and WorldPop on the 10 km grid scale was as high as 0.93, and the population distribution of SDpop2000 was obviously more accurate than that of WorldPop in the central Shandong, southwestern Taian, southern Jining, southern Linyi, northern Zaozhuang, and the northern Shandong coastal areas. In addition, the SDpop2000 better described the population distribution trend in Shandong Province, which was denser in western Shandong and the plains of northern Shandong than in the mountainous hilly areas in the central and southern Shandong, the coast of northern Shandong, and the hilly area of Shandong Peninsula. Overall, the spatialization method of township-level population data developed in this study significantly improved spatialization precision and is suitable for township-level population spatialization.
As the largest developing country in the world, China has witnessed rapid urbanization in the recent years. A large amount of natural land surface has been transformed into artificial land surface, leading to a series of environmental problems, among which the most prominent is urban heat island.Therefore, how to mitigate the urban heat island effect caused by the acceleration of urbanization process has become a hot research direction. To accurately analyze the influence of urban spatial pattern on thermal concentration, this article used two periods of remote sensing imagery, Landsat ETM+ on May 4, 2000 and Landsat OLI on July 27, 2016, to obtain the land cover information of Fuzhou and verified the accuracy. The hot spot analysis of the retrieved land surface temperature (LST) and impervious surface area (ISA) of Fuzhou were used to study the change characteristics, spatial concentration characteristics, and scale effect of LST in the past sixteen years of urbanization. The hot spot results show the following two findings. (1) The spatial thermal concentration could be better illustrated through analyzing the relationship between the distance from city center and LST. In 2000, the urban heat island effect was significant within a 1.03 km distance from the city center; however, in 2016 the distance increased to 2.1 km and the number of hot spots increased from three to five. During this period, the hot concentrated area (the hot and less hot areas) also increased from 15.7% to 47.3%. (2) Compared with other spatial autocorrelation analysis methods, Getis-Ord Gi* can help more directly analyze the impact of land cover change on LST and understand the details of the change of urban internal thermal intensity, because the formation of hot and cold spots depends on not only the level of LST. The hot spot method adopted in this study can be used for urban environmental protection and planning, and can also be used as a basis for urban land planning and thermal environmental impact analysis in the future. Meanwhile, the hot spot chart can be used to simulate urban microclimate and estimate the cooling effect of urban green space. In addition, comparative analyses of more multi-temporal and different cities can be further discussed in the future, especially studies on different types of cities, such as strip cities, polycentric cities and central cities.
Accurate and timely access to crop planting information is important for agricultural production management and national food security. Increased access to free satellite data, has allowed for further crop classification and physiological parameter inversion. Launched in June 2015, the Sentinel-2 satellite provides 13 spectral bands with high temporal, spatial and spectral resolution. This provides an opportunity to improve the operational accuracy and efficiency of rapid extraction of crop characteristics and large-scale crop planting area. The Sentinel-2 satellite provides free images at a high spatial resolution (10~30 m). Therefore, it is now possible to develop the next generation of agricultural product at both national and regional levels. Time series vegetation indices (VIs) are derived from this satellite data and are widely used to identify crop characteristics for classification. The Jianghan Plain (located in the middle and lower reaches of Yangtze River) was used as a case study area to evaluate and analyze the effect of the Sentinel-2 satellite images on rape planting area, based on an object-oriented method and differences in crop phenology. Firstly, the best estimate of rapeseed planting area was identified by using crop spectral information and normalized ve getation index at different growth stages. Secondly, a decision tree was formulated according to the difference in object features after multi-scale segmentation. Finally, the rapeseed planting area was calculated after the removal of non-vegetation areas, forests, winter wheat crops and other interferences based on the decision tree. It was found that image segmentation based on the Sentinel-2 images could identify various crop types effectively. The flowering characteristic of rapeseed is a key factor in distinguishing it from other crops. The difference in object features can effectively eliminate the effects of other land-use types on rapeseed classification and improve the accuracy and efficiency of crop classification. Results showed that the greatest classification contributors to the decision tree for distinguishing rapeseed were the normalized vegetation index (NDVI), near infrared (NIR), and brightness (Brightness). A total of 162 rapeseed sample points were used to calculate the confusion matrix. The overall classification accuracy of the rapeseed planting area was more than 98%, and the Kappa coefficient was 0.95. This shows that the Sentinel-2 data, combined with phenology information has great potential to extract large-scale crop planting area data, and can improve the accuracy and efficiency of calculating rape planting areas at both national and regional scales.
Remote Sensing technologies have been widely used in the investigation and dynamics monitoring of mangrove forests. However, problems remained that severely hinder the precise description and deep understanding of mangrove forests' dynamics. The problems include difficulties in remotely sensed data acquisition, the heavy workload of data preprocessing, and the lengthy time period in long time series monitoring. Based on Google Earth Engine (GEE), a cloud platform of remotely sensed data processing, this study used raw images of Landsat series satellites to produce an inter-annual mostly-cloudless (cloud coverage less than 5%) image collection of top-of-atmosphere reflectance (TOA). Then, classification rules were established based on three infrared-band TOAs (NIR, band near infrared; SWIR1, band shortwave infrared 1; SWIR2, band shortwave infrared 2) and three indices (NDVI, normalized difference vegetation index; NDWI, normalized difference water index; NDMI, normalized difference moisture index). Next, four land cover types, i.e., mangrove, mangrove-shrimp pond, impervious surface-bare land, and water body, were classified for mapping our case study area of Ngoc Hien, Vietnam from 1993 to 2017. Finally, the inter-annual land cover maps were used to analyze the characteristics of mangrove dynamics. The results showed that the long time-series inter-annual change monitoring of mangroves in cloudy and rainy regions can be implemented satisfactorily on the GEE platform. The image classification had an overall accuracy of over 80% for 86% of the study years, indicating that our proposed thresholds-based approach can effectively extract mangroves and mangrove-shrimp ponds. Through the analysis of inter-annual changes, the change process of mangroves in this region was depicted in details: it first increased, then decreased, and later, increased again. The correlation between the area changes of mangroves and mangrove-shrimp ponds was accurately detected to be negative. The inter-annual change monitoring of mangroves reduces the uncertainty of researching mangrove evolution processes, and quantifies in more details the conversions between mangroves and other land cover types. In so doing, the impacts of economic development, policies, and other factors on mangrove dynamics can then be assessed.
Cropping pattern is a fundamental aspect of land use. Obtaining accurate and timely crop spatial distribution information is very important to guide agricultural production, rational allocation of resources, and help solve the problem of food security. NDVI (Normalized Difference Vegetation Index) time series have been widely used in collecting vegetation information. Identification and information extraction of different crops can be effectively achieved by analyzing the growth period of crops and the time-series spectrum characteristics of NDVI. Most existing studies on NDVI time series are limited to moderate or low resolution remote sensing imagery, which affects the accuracy of extracting crop spatial distribution information. With the successful launch of the satellite Sentinel-2A, more opportunities have emerged for the construction of NDVI time series with high temporal and spatial resolution. In this paper, by use of typical phase Sentinel-2A imagery for Beian city, a NDVI dataset with a spatial resolution of 10 m covering the whole growth period of April-October was generated based on the Savitaky-Golay filtering smoothing method, and crop identification was implemented based on decision tree model and object-oriented classification. By analyzing NDVI time series curves of crop samples, we concluded that NDVI time series was able to clearly distinguish crop phenological differences and capture crop specific features in the study area. Furthermore, we also discussed the classification accuracy based on the typical phase data by the methods of object-oriented classification and support vector machine. Taking the field sample survey datum as true value, we analyzed the results of the two classification methods. The results show that the processed NDVI time series with high resolution over the entire crop growth cycle represent different crop phenological characteristics appropriately. It is able to reflect the crops growth condition accurately and distinguish different crops effectively. The decision tree model integrated with the object-oriented classification method had the highest accuracy in crop classification as compared to other classification methods, with its overall accuracy and kappa coefficient being 96.2% and 0.892, respectively. This research show that Sentinel-2A NDVI with high resolution can be used for crop mapping, and can be applied to crop classification over large areas, thanks to Sentinel-2A imagery's wide coverage. Furthermore, the Savitzky Golay (S-G) method was used for NDVI time series smoothing, and the results indicate that the processed NDVI time series can better represent crop phenological characteristics. Then the decision tree model integrated with the object-oriented classification method was used for crop classification based on typical phase multi-spectral imagery and the smoothed NDVI time series, which improved the overall accuracy by 7.7% and the kappa coefficient by 0.055. The approach proposed in this paper provides important reference for crop mapping in other agricultural regions.
Jiangsu province, with 13 municipalities and located in the east of China, is an important part of the Yangtze river delta economy belt. The temperature is appropriate and the rainfall is moderate. Jiangsu province enjoys a moderate climate, which is suitable for the agricultural development. Winter wheat is distributed throughout the whole province, whereas the planting structure of winter rapeseed is complex and mainly scattered in Southern Jiangsu. As reported by the State Statistics Bureau, the total planting area of winter wheat and winter rapeseed in Jiangsu ranked the fifth and seventh in China, respectively, during the last 10 years. Fast obtaining the precise planting area of these two crops in Jiangsu is crucial for the agricultural development. Remote sensing classification based on local host can obtain spatial distribution of crops with high accuracy, but is time-consuming. With the development of geographical big data, cloud platform, and cloud computation, the Google Earth Engine (GEE), a global scale geospatial analysis platform based on the cloud platform, has brought new opportunities for rapid remote sensing classification. Based on the GEE cloud platform, a time-saving method of obtaining the spatial distribution of winter wheat and winter rapeseed by use of sentinel-2 data in Jiangsu was proposed. First, 119 sentinel-2 images without cloud were obtained using the GEE in Jiangsu. The time interval was set from March 1 to June 1, 2017, and the space area was Jiangsu province. Based on the spatio-temporal information, the 119 remote sensing images were mosaicked and clipped. Secondly, remote sensing indices, texture, and terrain features were calculated respectively, and the original features were extracted. The original feature space was optimized by an algorithm named Separability and Thresholds (SEaTH algorithm). Finally, four classifiers including naive Bayes, support vector machine, classification regression tree, and random forest were tested and evaluated by the average assessment accuracy. The spatial distribution information of winter wheat and winter rapeseed were obtained quickly. The following conclusions are drawn: (1) the GEE can quickly complete pre-processing of cloud-masking, image-mosaicking, image-clipping, and feature extraction, which is superior to the local processing. (2) The distance values of J-M that are higher than 1 and rank top two highest can reduce the number of features from 28 to 11 and effectively compress the original feature space. (3) With the combined training of spectral, texture and terrain features, the average assessment accuracy of naive Bayes, support vector machine, classification regression tree, and random forest was 61%, 87%, 89% and 92%, respectively.
Vegetation is a sensitive indicator of global climatic changes, and hydrothermal conditions are the main abiotic factors that determine the phenology, spatial pattern, and dynamics of vegetation. Thus, against the background of a changing climate, the climate-vegetation relationship is a hot topic in current global change research. Using geodetector, this study integrated climatic factors (e.g., average temperature, precipitation, water vapor pressure, humidity, sunshine hours, standardized precipitation evapotranspiration index), topographic factors (e.g., slope and elevation), and anthropogenic factors to determine the dominant factors that influenced the normalized difference vegetation index (NDVI) in the Beijing-Tianjin-Hebei region from 2006 to 2015. Different seasons, geomorphological types, and vegetation types were considered during the quantitative attribution analysis. This study revealed the temporal and spatial pattern of vegetation, and the response of vegetation to climate and non-climate factors over the past 10 years, and provided a reference for the construction and restoration of ecological engineering. Trend analysis showed that the NDVI increased during this period, albeit with differences on different spatial scales. In montane regions, the NDVI increased more rapidly than in plains, terraces, and hills. In different vegetation-covered areas, the NDVI increased most rapidly in broadleaf forest, followed by shrubland and coniferous forest. Based on the results of the quantitative distribution analysis, at the temporal scale of one year, precipitation was the dominant factor driving NDVI and explained 39.4% of the spatial distribution, while the interaction of rainfall and land use was the dominant interaction factor, with a q value of 0.582. We observed seasonal and regional differences in the response of NDVI to climatic factors. In the four seasons, vapour pressure was the dominant factor for the spatial distribution of NDVI; humidity is the dominant factor in summer and autumn; and in winter, land use was the dominant factor for NDVI distribution. The explanatory power of the influencing factors on NDVI in the growing season differed in diverse geomorphological types. In montane areas, with increasing elevation, the q value of average temperature decreased. The explanatory power of impacting factors on NDVI of the growing season was different among diverse vegetation types. For different vegetation types, the explanatory power of precipitation on the spatial distribution of NDVI in the growing season was greater than that of mean temperature, with the q value ranked as following grassland > cultivated vegetation > shrubland > broadleaf forest >coniferous forest. In coniferous forest distributed areas, the explanatory power of single factors was insignificant; however, the interaction between two factors can greatly enhance the q value, and the interaction between moisture factors and topographic factors was the dominant factor.
Sri Lanka is an important node on the Maritime Silk Road, where rainfall is abundant in quantity but uneven in terms of spatiotemporal distribution. There is obvious seasonal water shortage. Monitoring the changes of water cover area in inland lakes and reservoirs is important for guiding the development and utilization of water resources. To understand the spatial distribution characteristics and temporal variations of lakes and reservoirs in Sri Lanka, this paper, based on Landsat series imagery, analyzed and compared the precision of different water extraction models on the images, following which the optimal algorithm was determined. A typical reservoir was chosen to analyze the interannual and monthly variations of the water cover sizes. The optimal water extraction algorithm was applied to the inland lakes and reservoirs in 1995, 2005, and 2015. Lakes and reservoirs were divided into four grades by area. The number and area of lakes and reservoirs of different grades in each year were counted, and their spatiotemporal variation characteristics were examined. Conclusions can be made according to the results as the following statements: (1) The water body extraction model based on the Normalized Difference Water Index (NDWI) with threshold value from the Otsu method (OTSU) had the best accuracy and was suitable for the water body extraction in Sri Lanka. The overall classification accuracy is above 97% and it has the lowest mis-extraction rate and the missing rate. (2) The water cover area of typical reservoir showed a fluctuatingly increase trend in Augusts from 1988 to 2018. The smallest water cover area occurred in 1992, and the largest was in 2013. The water cover area of reservoir was also of large intra-annual fluctuations. In 2017, the biggest water cover area appeared in February, while the smallest appeared in September, with a discrepancy of 2.24 times between the cover area in February and September, exactly the ends of local rainy season and dry season, respectively. (3) From 1995 to 2015, the number and area of lakes and reservoirs of different grades increased to some extent, and the trend of lake and reservoir water resources was increasing. Findings of the research will provide necessary data support for the management and planning of soil and water resources in Sri Lanka.
The methodology of combining sampling-based ground survey and satellite imagery classification has been widely used in estimating crop acreage on large scales. Use of unmanned aerial vehicle (UAV) imagery has a series of merits including low cost, high efficiency, and high resolution, which make it possible to quickly monitor the agricultural conditions over a specific area. With a research focus on rice sample plots, this study used a portable UAV Mavic Pro to obtain aerial imagery. The UAV imagery were preprocessed to generate an orthophoto with a resolution of 3.95 cm/pix. By adopting the object-oriented classification philosophy, visual assessment, and the Estimation of Scale Parameter (ESP) tool, the optimal segmentation scale was determined to be 300. The support vector machine, random forest, and nearest neighborhood classifiers were employed and contrasted for imagery classification and the extraction of rice acreage; visual interpretation was used for assessing the accuracy of the classification results. The best automatic classification method turned out to be nearest neighborhood classification, with its user accuracy of rice being 95% and the area consistency accuracy 99%. The findings show that use of UAV imagery and automatic classification can quickly acquire high-resolution imagery and extract rice acreage in rice growing areas on plains. Moreover, high-resolution UAV imagery can be used as ground truth data when cropland is in shadow. The proposed approach helps provide validation samples for estimating rice acreage and production on large scales.
