Fast R-CNN深度学习和无人机遥感相结合在松材线虫病监测中的初步应用研究

松材线虫病因其破坏性强、传播速度快和防治难度大等特点,严重威胁着我国的松林资源.及时发现、定位和清理病死松树是控制松材线虫病蔓延的有效手段.本研究利用小型无人机获得松材线虫病疫点的可见光和多光谱的航摄影像.根据松树针叶颜色变化,将松材线虫Bursaphelenchus xylophilus侵染的松树分为病树和枯死树两种类型.将无人机遥感正摄影像图切割成瓦片图,根据不同植被指数的特征差异,筛选出含病树和枯死树的瓦片图.训练Fast R-CNN深度学习框架形成最终模型,通过模型运算获得病枯死松树的分布地图及坐标点位置.研究结果显示Fast R-CNN深度学习和无人机遥感相结合能有效识别出病树和枯死树,正确率分别达到90％和82％,漏检率分别为23％和34％,可为大面积监测松材线虫病的发生现状和流行动态、评估防控效果和灾害损失提供技术支撑.     

卫星遥感监测产品在中国森林生态系统的验证和不确定性分析——基于海量无人机激光雷达数据

准确获取森林结构参数对森林生态系统研究及其保护有着重要意义。卫星遥感数据作为获取大尺度森林结构参数的重要数据源, 已被制作成各种植被监测产品并被应用于森林质量状况变化评估、森林生物量估算以及森林干扰和生物多样性监测等研究。然而, 这些卫星遥感植被监测产品针对中国复杂多样的森林区域缺乏有效验证, 在不同林况和地形条件下的不确定性也不明确。激光雷达具备高精度三维信息采集的优势, 在国内外已被广泛用于森林生态系统监测和卫星遥感产品验证。为此, 该研究利用在中国114个样地收集的153 km²的无人机激光雷达数据, 构建了我国森林结构参数验证数据集, 并以此为基础对3套全球遥感监测产品(全球叶面积指数(GLASS LAI)、全球冠层覆盖度(GLCF TCC)、全球冠层高度(GFCH))进行了像元尺度的验证, 并分析了其在不同坡度、覆盖度和林型条件下的不确定性。研究结果表明: 与无人机激光雷达获取的叶面积指数、覆盖度以及冠层高度相比, GLASS LAI、GLCF TCC、GFCH在中国森林区域均存在一定的不确定性, 且受林况和地形因素影响的程度不一致。对GLASS LAI和GLCF TCC影响的最大因素分别为林型和覆盖度; 而GFCH则更易受地形坡度和覆盖度的影响。     

遥感在林冠动态监测研究中的应用

 林冠动态大致包括三方面的内容,即由病虫害、林火等引起的林冠变化、由大风等灾害引起的林隙动态、以及树冠和林冠的正常变化等。遥感在林冠动态研究中的地位和作用已被广泛认知,国内外在此方面的研究已积累了丰富经验。进行林冠动态研究所利用的遥感数据主要有Landsat TM卫     

遥感技术在昆虫生态学中的应用途径与进展

结合雷达、航空和卫星遥感本身的特点 ,从害虫本身、害虫所造成的危害、影响害虫发展的环境因子三方面介绍了遥感技术在区域性害虫的早期监测及预测中的应用途径与最新进展。针对害虫本身的个体大小、可动性和种群所在空间尺度的影响 ,作者强调应发挥不同遥感系统各自独特的优势 ,同时 ,综合应用“3S”技术和时空模型方法 ,才能够实现害虫动态的可视化、立体化、实时化和精确化监测及预测。     

An algorithm for early rice area mapping from satellite remote sensing data in southwestern Guangdong in China based on feature optimization and random Forest

Estimating of rice areas using images obtained from satellite remote sensing is important for guiding operators. The object of this study was the Sentinel-2A/B image data of the rice planting demonstration regions in southwestern Guangdong, China. We designed an algorithm for early rice area mapping based on feature optimization and random forest (RF). For modeling, we selected 35 common remote sensing features and applied out-of-bag (OOB) to construct 7 feature combinations. The results showed that the overall accuracy (OA) and Kappa coefficient of the RF with the best combination were 91.23% and 87.55%, respectively. Compared with support vector machine (SVM) and back propagation neural network (BPNN), the model result of RF was also the best among the three. Additionally, the maximum error of the rice area was less than 16% when the model was transferred to other regions in Guangdong. The feature optimization and RF-based algorithm proposed in this study can effectively map the early rice region. It can be applied to estimate rice area based on satellite remote sensing image data and reveal the ecological status of rice cultivation in southwestern Guangdong.     

Monitoring <i>Thosea sinensis</i> Walker in Tea Plantations Based on UAV Multi-Spectral Image

Thosea sinensis Walker (TSW) rapidly spreads and severely damages the tea plants. Therefore, finding a reliable operational method for identifying the TSW-damaged areas via remote sensing has been a focus of a research community. Such methods also enable us to calculate the precise application of pesticides and prevent the subsequent spread of the pests. In this work, based on the unmanned aerial vehicle (UAV) platform, five band images of multispectral red-edge camera were obtained and used for monitoring the TSW in tea plantations. By combining the minimum redundancy maximum relevance (mRMR) with the selected spectral features, a comprehensive spectral selection strategy was proposed. Then, based on the selected spectral features, three classic machine learning algorithms, including random forest (RF), support vector machine (SVM), and k-nearest neighbors (KNN) were used to construct the pest monitoring model and were evaluated and compared. The results showed that the strategy proposed in this work obtained ideal monitoring accuracy by only using the combination of a few optimized features (2 or 4). In order to differentiate the healthy and TSW-damaged areas (2-class model), the monitoring accuracies of all the three models were computed, which were above 96%. The RF model used the least number of features, including only SAVI and Band_red. In order to further discriminate the pest incidence levels (3-class model), the monitoring accuracies of all the three models were computed, which were above 80%, among which the RF algorithm based on SAVI, Band_red, VARI_green, and Band_{red_edge} features achieve the highest accuracy (OAA of 87%, and Kappa of 0.79). Considering the computational cost and model accuracy, this work recommends the RF model based on a few optimal feature combinations to monitor and distinguish the severity of TSW in tea plantations. According to the UAV remote sensing mapping results, the TSW infestation exhibited an aggregated distribution pattern. The spatial information of occurrence and severity can offer effective guidance for precise control of the pest. In addition, the relevant methods provide a reference for monitoring other leaf-eating pests, effectively improving the management level of plant protection in tea plantations, and guaranting the yield and quality of tea plantations.     

Detection and classification of diseased pine trees with different levels of severity from UAV remote sensing images

The accurate detection and classification of diseased pine trees with different levels of severity is important in terms of monitoring the growth of these trees and for preventing and controlling disease within pine forests. Our method combines a DDYOLOv5 with a ResNet50 network for detecting and classifying levels of pine tree disease from remote sensing UAV images. In this approach, images are preprocessed to increase the background diversity of the training samples, and efficient channel attention (ECA) and hybrid dilated convolution (HDC) modules are introduced to DDYOLOv5 to improve the detection accuracy. The ECA modules enable the network to focus on the characteristics of diseased pine trees, and solve the problem of low detection accuracy caused by the similarities in color and texture between diseased pine trees and the complex backgrounds. The HDC modules capture the contextual information of targets at different scales; they increase the receptive field to focus on targets of different sizes, and address the difficulty of detection caused by large variations in the shapes and sizes of diseased pine trees. In addition, a low confidence threshold is adopted to reduce missed detections and a ResNet50 classification network is applied to classify the detection results into different levels of severity, in order to reduce the number of false detections and improve the classification accuracy. Our experimental results show that the proposed method improves the precision by 13.55%, the recall by 5.06% and the F1-score by 9.71% on 8 test images compared with YOLOv5. Moreover, the detection and classification results from our approach show that it outperforms classical deep learning object detection methods such as Faster R-CNN and RetinaNet.     

Satellite remote sensing of wetlands

To conserve and manage wetland resources, it is important to inventoryand monitor wetlands and their adjacent uplands. Satellite remote sensing hasseveral advantages for monitoring wetland resources, especially for largegeographic areas. This review summarizes the literature on satellite remotesensing of wetlands, including what classification techniques were mostsuccessful in identifying wetlands and separating them from other land covertypes. All types of wetlands have been studied with satellite remote sensing.Landsat MSS, Landsat TM, and SPOT are the major satellite systems that have beenused to study wetlands; other systems are NOAA AVHRR, IRS-1B LISS-II and radarsystems, including JERS-1, ERS-1 and RADARSAT. Early work with satellite imageryused visual interpretation for classification. The most commonly used computerclassification method to map wetlands is unsupervised classification orclustering. Maximum likelihood is the most common supervised classificationmethod. Wetland classification is difficult because of spectral confusion withother landcover classes and among different types of wetlands. However,multi-temporal data usually improves the classification of wetlands, as doesancillary data such as soil data, elevation or topography data. Classifiedsatellite imagery and maps derived from aerial photography have been comparedwith the conclusion that they offer different but complimentary information.Change detection studies have taken advantage of the repeat coverage andarchival data available with satellite remote sensing. Detailed wetland maps canbe updated using satellite imagery. Given the spatial resolution of satelliteremote sensing systems, fuzzy classification, subpixel classification, spectralmixture analysis, and mixtures estimation may provide more detailed informationon wetlands. A layered, hybrid or rule-based approach may give better resultsthan more traditional methods. The combination of radar and optical data providethe most promise for improving wetland classification.     

Remote sensing of upland vegetation: the potential of high spatial resolution satellite sensors

Aim Traditional methodologies of mapping vegetation, as carried out by ecologists, consist primarily of field surveying or mapping from aerial photography. Previous applications of satellite imagery for this task (e.g. Landsat TM and SPOT HRV) have been unsuccessful, as such imagery proved to have insufficient spatial resolution for mapping vegetation. This paper reports on a study to assess the capabilities of the recently launched remote sensing satellite sensor Ikonos, with improved capabilities, for mapping and monitoring upland vegetation using traditional image classification methods. Location The location is Northumberland National Park, UK. Methods Traditional remote sensing classification methodologies were applied to the Ikonos data and the outputs compared to ground data sets. This enabled an assessment of the value of the improved spatial resolution of satellite imagery for mapping upland vegetation. Post‐classification methods were applied to remove noise and misclassified pixels and to create maps that were more in keeping with the information requirements of the NNPA for current management processes. Results The approach adopted herein for quick and inexpensive land cover mapping was found to be capable of higher accuracy than achieved with previous approaches, highlighting the benefits of remote sensing for providing land cover maps. Main conclusions Ikonos imagery proved to be a useful tool for mapping upland vegetation across large areas and at fine spatial resolution, providing accuracies comparable to traditional mapping methods of ground surveys and aerial photography.     

我国林业重大害虫松毛虫的灾害研究进展

我国松毛虫种类丰富,已报道27种,其中6种频繁暴发,年危害松林面积达200万hm2以上、经济损失数亿元。松毛虫灾害研究可以为综合治理松毛虫提供重要的理论依据与相关技术。近年来,我国在影响松毛虫种群发生的关键性因子分析、性信息素的研究与开发利用、松树挥发性物质的成分鉴定及其对寄生蜂或寄生蝇的定位作用、卫星遥感监测技术等方面的研究取得了较大进步,但缺少系统性的归纳总结。为此,文章就近些年来我国在松毛虫灾害机制与暴发机理研究、治理现状与技术措施等方面进行综述,展望今后我国松毛虫治理研究工作的重要方向和趋势。     

松材线虫病发生点格局及影响因素

松材线虫病在中国大陆造成了巨大的生态与经济价值损失,南方地区尤为严重,分析松材线虫病空间分布、量化环境因素对其发生的影响对于松材线虫病的防控整治具有重要意义。本研究以江西省赣州市南康区松材线虫病为研究对象,采用核平滑密度、Ripley’s K函数、点过程模拟等空间点格局分析方法,探讨了区域松材线虫病发生的空间格局及其对环境变量的响应。结果表明: 研究区松材线虫病的发生不是随机分布,而是存在显著的空间聚集区。地形因子、植被因子和人类活动因子是影响松材线虫病空间异质性分布的主要因素。空间点格局分析表明,海拔、坡度、距最近道路距离、道路密度、距最近居民点距离、郁闭度和植被类型对松材线虫病的发生具有重要影响。在森林病害管理中,除了加强因人类活动引起病害传播源的管控外,还应该考虑地形、植被类型等特征进行松材线虫病害的综合预警监测。     

A systematic review on hyperspectral imaging technology with a machine and deep learning methodology for agricultural applications

The globe's population is increasing day by day, which causes the severe problem of organic food for everyone. Farmers are becoming progressively conscious of the need to control numerous essential factors such as crop health, water or fertilizer use, and harmful diseases in the field. However, it is challenging to monitor agricultural activities. Therefore, precision agriculture is an important decision support system for food production and decision-making. Several methods and approaches have been used to support precision agricultural practices. The present study performs a systematic literature review on hyperspectral imaging technology and the most advanced deep learning and machine learning algorithm used in agriculture applications to extract and synthesize the significant datasets and algorithms. We reviewed legal studies carefully, highlighted hyperspectral datasets, focused on the most methods used for hyperspectral applications in agricultural sectors, and gained insight into the critical problems and challenges in the hyperspectral data processing. According to our study, it has been found that the Hyperion hyperspectral, Landsat-8, and Sentinel 2 multispectral datasets were mainly used for agricultural applications. The most applied machine learning method was support vector machine and random forest. In addition, the deep learning-based Convolutional Neural Networks (CNN) model is mainly used for crop classification due to its high performance with hyperspectral datasets. The present review will be helpful to the new researchers working in the field of hyperspectral remote sensing for agricultural applications with a machine and deep learning methods.     

东亚地区夏季干旱、强台风事件与松树枯萎病的关系

松树枯萎病的发生和流行给东亚的日本、中国和韩国带来了不小的损失,引起许多国家的重视. 本文应用实地观测和气象数据分析等方法,研究了东亚地区松树枯萎病的发生和流行特点,以及与灾害气象事件之间的关系.结果表明: 在日本、中国和韩国,持续的夏季干旱少雨和强台风等极端气象事件能够诱发松树枯萎.在极端干热的环境中,松树常出现能量代谢失调, 以至于受胁迫的松树整株枯萎;而在低温多雨的年份,松树枯萎少,甚至没有发生枯死现象.在松材线虫及其媒介昆虫侵染之前,松树的活力业已下降.松树枯萎病似乎应该局限在台风频发和持续干热的地区.在自然环境优越、少有台风和干热事件出现且没有不当扩大松树栽培范围的地区,松树枯萎病大面积发生的可能性不大.     

基于高光谱特征的雅氏落叶松尺蠖虫口密度估算

[目的]多年来,蒙古高原典型落叶松害虫雅氏落叶松尺蠖Eeannis jacobssoni发生频繁,使森林生态系统遭到严重破坏.虫口密度可直接描述森林虫害严重程度,进而及时、快速获得害虫虫口密度信息显得极为重要.本研究旨在依据雅氏落叶松尺蠖暴发区的落叶松光谱实测数据和虫口密度数据,构建基于高光谱特征的虫口密度估算方法.[...     

Response of soil microbial communities to changes in a forest ecosystem brought about by pine wilt disease

Japan has suffered a lot from forestry losses due to pine wilt disease caused by pinewood nematode infestations. Studies were conducted regarding its causative agent and the effects of natural vegetation succession after pine wilt disease, but its effects on microorganisms were not given equal attention. This study determined the effects of pine wilt disease on light conditions, soil microbial biomass, litter decomposition, microbial abundance and the physical and chemical properties of the soil. Results showed that in a forest currently affected by pine wilt disease, there was higher light penetration, greater microbial biomass carbon, and a faster rate of litter decomposition. Microbial abundance was shown to be reduced in pine wilt affected areas. There were close correlations between the biological and physicochemical properties of the soil, but the reason for the decrease in microbial abundance is not yet well understood, and thus requires further study.     

Development of secondary pine forests after pine wilt disease in western Japan

Abstract. The development of secondary Pinus densiflora (Japanese red pine) forests after pine wilt disease was studied through phytosociological analysis, estimation of forest structure before disease and size-structure, tree ring and stem analyses. Following the end of the disease, the growth of previously suppressed small oak trees was accelerated. This is quite different from the development of forests following fire, which starts with the establishment of pine seedlings. Pine wilt disease shifted the dominance of secondary forests from Pinus densiflora to Quercus serrata oak forest. In pine forests, disturbance by fire is important for forest maintenance. In contrast, disturbance by pine wilt disease leads to an acceleration of succession from pine forest to oak forest.     

松树萎蔫的成因与控制策略

已报道的松树萎蔫枯死类型有多种,它们之间存在内在的联系,其发生及严重与否涉及多种生态因子,是立地环境、感病松树、媒介昆虫、病原线虫、病原微生物综合作用的结果,基于生态学的可持续有害生物管理应是防治松树萎蔫的有效途径。     

Oil palm trees detection and counting on Microsoft Bing Maps Very High Resolution (VHR) satellite imagery and Unmanned Aerial Vehicles (UAV) data using image processing thresholding approach

Palm oil is one of the highest producing vegetable oil crops globally, with production increasing rapidly over the last 40 years from 5 million tonnes in 1980 to 74.5 million tonnes in 2019. This increase in production is in line with the high demand for vegetable oils worldwide. The accurate monitoring and statistics of oil palm plantation data are essential to support effective and efficient decision-making. However, the most commonly adopted data collection still uses conventional methods, i.e., field surveys, which are highly dependent on the massive amount of human resources, cost, processing time, and difficulty reaching remote areas. Remote sensing using satellite and UAV imagery can be an alternative in data collection due to its distinct advantages with a more efficient labor force, affordable cost, shorter time updates, and covering areas that are difficult to reach. In this work, we investigate the utilization of remote sensing data from Microsoft Bing Maps Very High Resolution (VHR) satellite imagery and Unmanned Aerial Vehicle (UAV) data using the image processing thresholding method for detecting and counting oil palm trees. The combination of Hue, Saturation, and Value (HSV) conversion, the Otsu segmentation thresholding, and contours detection and counting methods are used in our approach to enhance the accuracy of captured features of oil palm trees. The detection results are further categorized into best-case, average-case, and worst-case detection to comprehend the challenging real-world situation, based on the quality of captured imageries and model prediction results. Our proposed approach achieves better and more promising results when using UAV image data. This is indicated by an average true-positive rate (TPR) of 88 to 97% in best-case data input, 70 to 81% in average-case data input, 30 to 53% in worst-case data input, and 10% in estimation error. Meanwhile, Bing image data provides an average true-positive rate (TPR) of 9 to 13% in the best-case data input, 5 to 6% in the average-case data input, and 1 to 3% in the worst-case data input, with an average estimation error up to 78%. Overall, our proposed approach can perform the oil palm trees detection and counting. We also suggest that higher resolution imagery than Microsoft Bing maps or single utilization of UAV data can be considered input for better results. Our study could further be beneficial in providing more scalable and accurate plantation statistics.     

Succession of secondary forests affected by pine wilt disease in western Japan followed on vegetation maps

Abstract. Succession, changes in the distribution pattern of forest vegetation, and Pinus forest survival following pine wilt disease were clarified based on phytosociological analysis and vegetation maps. Survival of Pinus forests was restricted to the early successional stages, which were located on ridges and the upper part of slopes. Subsequent to pine wilt disease, the succession progressed from early to late substages of Pinus forest, mixed deciduous and evergreen Quercus, to evergreen Quercus forest. Succession occurs in abandoned pine forests which apparently are in a bad state and are vulnerable to attacks by pine wilt disease.     

基于多源遥感数据的面向对象林分类型识别

林分类型的识别是森林资源监测的核心问题之一.为研究多源遥感数据协同的面向对象林分类型分类识别,采用Radarsat-2数据和QuickBird遥感影像协同进行面向对象分类.在面向对象分类过程中,采用3种分割方案:单独使用QuickBird遥感影像分割;单独使用Radarsat-2数据分割;Radarsat-2&QuickBird协同分割.3种分割方案均采用10种分割尺度(25～250,步长25),应用修正的欧式距离3指标评价不同分割方案的分割结果,确定最优分割方案及最优分割尺度.在最优分割结果的基础上,基于地形、高度、光谱及共同特征的不同特征组合,应用带有径向基(RBF)核函数的支持向量机(SVM)分类器进行杉木林、马尾松林、阔叶林3种林分类型识别.结果表明:与单独使用一种数据相比,Radarsat-2数据和QuickBird遥感影像协同方案在面向对象林分类型分类方面具有优势.Radarsat-2&QuickBird协同分割方案,以最优尺度参数100进行分割时,分割结果最好.在最优分割结果的基础上,应用两种数据源提取的全部特征进行面向对象林分类型识别的精度最高(总精度为86%,Kappa值为0.86).本研究结果不仅可为多源遥感数据结合进行林分类型识别提供参考和借鉴,而且对于森林资源调查和监测有现实意义.