三江源大型食草动物数量无人机自动监测方法

应用无人机进行动物调查的研究越来越多,但是缺乏有效的自动处理图像方法,使得目视解释的工作量较大。为研究快速与准确估算三江源区域大型食草动物数量的方法,本文使用两种无人机进行7个站点的遥感影像采集。首先对无人机影像进行灰度化,使矩阵维数下降但梯度信息仍然保留,使运算速度大幅度提高;其次对影像开展高斯滤波,高斯滤波将数据进行能量转化,排除掉属于低能量部分的噪声;第三开展阈值处理得到二值化图像;采用样本中动物形态学特征开展形态运算,先用开运算消除小物体尽可能排除干扰,同时不误删牲畜,再用闭运算排除小型黑洞将同一对象连通不重复计数;最后从二值图像中检索轮廓,并返回检测到的轮廓的个数;从而自动获得主要大型食草动物物种数量与空间分布。精度检验方法为手工计数与自动计数结果比较,相对误差3.1%~6.5%,大多数情况均可达到此水平。采用计算机自动处理图像后,每张图像处理和计数的平均时间小于3 s。无人机影像的自动处理方法可为今后大规模进行藏羊、牦牛、西藏野驴和藏原羚等动物调查提供一种有效、可靠的技术途径。     

Evaluating new technology for biodiversity monitoring: Are drone surveys biased?

Drones and machine learning‐based automated detection methods are being used by ecologists to conduct wildlife surveys with increasing frequency. When traditional survey methods have been evaluated, a range of factors have been found to influence detection probabilities, including individual differences among conspecific animals, which can thus introduce biases into survey counts. There has been no such evaluation of drone‐based surveys using automated detection in a natural setting. This is important to establish since any biases in counts made using these methods will need to be accounted for, to provide accurate data and improve decision‐making for threatened species. In this study, a rare opportunity to survey a ground‐truthed, individually marked population of 48 koalas in their natural habitat allowed for direct comparison of the factors impacting detection probability in both ground observation and drone surveys with manual and automated detection. We found that sex and host tree preferences impacted detection in ground surveys and in manual analysis of drone imagery with female koalas likely to be under‐represented, and koalas higher in taller trees detected less frequently when present. Tree species composition of a forest stand also impacted on detections. In contrast, none of these factors impacted on automated detection. This suggests that the combination of drone‐captured imagery and machine learning does not suffer from the same biases that affect conventional ground surveys. This provides further evidence that drones and machine learning are promising tools for gathering reliable detection data to better inform the management of threatened populations.     

Hot stuff in the bushes: Thermal imagers and the detection of burrows in vegetated sites

1. Thermal imaging technology is a developing field in wildlife management. Most thermal imaging work in wildlife science has been limited to larger ungulates and surface‐dwelling mammals. Little work has been undertaken on the use of thermal imagers to detect fossorial animals and/or their burrows. Survey methods such as white‐light spotlighting can fail to detect the presence of burrows (and therefore the animals within), particularly in areas where vegetation obscures burrows. Thermal imagers offer an opportunity to detect the radiant heat from these burrows, and therefore the presence of the animal, particularly in vegetated areas. Thermal imaging technology has become increasingly available through the provision of smaller, more cost‐effective units. Their integration with drone technology provides opportunities for researchers and land managers to utilize this technology in their research/management practices.
2. We investigated the ability of both consumer (<AUD$20,000) and professional imagers (>AUD$65,000) mounted on drones to detect rabbit burrows (warrens) and entrances in the landscape as compared to visual assessment.
3. Thermal imagery and visual inspection detected active rabbit warrens when vegetation was scarce. The presence of vegetation was a significant factor in detecting entrances (p < .001, α = 0.05). The consumer imager did not detect as many warren entrances as either the professional imager or visual inspection (p = .009, α = 0.05). Active warren entrances obscured by vegetation could not be accurately identified on exported imagery from the consumer imager and several false‐positive detections occurred when reviewing this footage.
4. We suggest that the exportable frame rate (Hz) was the key factor in image quality and subsequent false‐positive detections. This feature should be considered when selecting imagers and suggest that a minimum export rate of 30 Hz is required. Thermal imagers are a useful additional tool to aid in identification of entrances for active warrens and professional imagers detected more warrens and entrances than either consumer imagers or visual inspection.

     

基于多色彩空间的YOLOv5松枯死树检测方法

【目的】针对在松枯死树监测实践中,从无人机航拍RGB影像中自动识别松枯死树漏检率高的问题,提出了一种生产应用场景下基于多色彩空间的YOLOv5松枯死树高精度自动识别新方法。【方法】利用无人机采集大面积松材线虫病发生林分的RGB图像,用Pix4Dmapper软件拼接,用LabelImg开源软件建立VOC格式的松枯死树数据集,分别用Faster R-CNN、YOLOv3、YOLOv4、YOLOv5、SSD和EfficientDet等6种基于深度学习的目标检测算法对数据集进行训练和测试,以精确率、召回率、平均准确率以及F₁分数作为评价指标筛选出最优目标检测算法;然后将采集的RGB图像转换成LAB和HSV色彩空间图像,再将这3个色彩空间的图像分别用最优目标检测算法进行训练,得到目标在每个色彩空间的边界框,使用非极大值抑制算法对这些边界框进行处理,得到最优边界框实现松枯死树自动识别。【结果】6种算法均取得良好效果,其中YOLOv5模型为最优算法,其精准率、平均查准率和F₁分数在6种算法中均最高,分别达到97.58%、82.40%和0.85。通过3个色彩空间融合后,反映漏检情况的召回率由74.54%提高到98.99%,平均准确率提升至98.39%。【结论】基于多色彩空间的YOLOv5模型能够显著提高从无人机航拍RGB影像中检测松枯死树的精度,为松枯死树监测提供了有力工具,也有助于松材线虫病的防治。     

Development of a PWM Precision Spraying Controller for Unmanned Aerial Vehicles

This paper presents a new Pulse Width Modulation (PWM) controller for Unmanned Aerial Vehicle (UAV) precisionsprayer for agriculture using a TL494 fixed-frequency pulse width modulator together with a data acquisition board and developedsoftware. An UAV can be remotely controlled or flown autonomously by pre-programmed flight plans. The PWMcontroller was implemented through the guidance system on the UAV with control commands sent between the UAV helicopterand the ground control station via a wireless telemetry system. The PWM controller was tested and validated using LabVIEW8.2. Several analyses were performed in a laboratory to test different control signals. The results show that the PWM controllerhas promise as a higher precision technique for spray applications, which will improve efficiency of pesticide application,especially in crop production areas.     

Applications of an unmanned aerial vehicle and thermal-imaging camera to study ducks nesting over water

Finding and monitoring nests are key components of avian research, but they are often expensive, time-consuming, and inefficient operations. This is certainly true for diving ducks that nest in wetlands with thick emergent vegetation where nests are typically located by teams of technicians that wade through a marsh and beat vegetation with sticks, hoping to flush incubating females or encounter nests without a female present. Taking advantage of recent advances in both unmanned aerial vehicles (UAVs) and thermal-imaging cameras, our objectives were to (1) compare our ability to locate duck nests using a UAV and using traditional on-foot searching methods, and (2) determine if nests monitored remotely with the UAV had different survival rates than nests monitored with traditional nest-site visits. We searched for nests with a UAV system in southern Manitoba during the springs of 2018 and 2019. Using the UAV, we located 48 nests not found by ground crews, ground crews found 164 nests not found with the UAV, and 71 nests were found using both methods. Overall, nests were less likely to be detected with the UAV (0.34) than by ground crews (0.71), but surveys were completed approximately four times faster with the UAV. Detectability of nests varied among duck species (range = 0.55–0.04). We found no difference in nest survival between nests monitored with the UAV (0.95) and those repeatedly visited by ground crews (0.95). However, in 2018, ground monitoring resulted in 19 nests being abandoned by females, compared to only one monitored with the UAV. Our results demonstrate that UAVs equipped with thermal cameras can be used to find nests of ducks located over water, with greater success for species that nest earlier and those whose nests are not buried under matted vegetation. Furthermore, monitoring nests with the UAV resulted in lower rates of nest abandonment, and survival of nests monitored with the UAV was similar to that of nests monitored using traditional methods. Additional species- and habitat-specific studies are needed to fully understand the utility and challenges associated with using UAVs equipped with thermal imaging to survey species of wetland wildlife.     

Aerial photogrammetry and tag-derived tissue density reveal patterns of lipid-store body condition of humpback whales on their feeding grounds

Monitoring the body condition of free-ranging marine mammals at different life-history stages is essential to understand their ecology as they must accumulate sufficient energy reserves for survival and reproduction. However, assessing body condition in free-ranging marine mammals is challenging. We cross-validated two independent approaches to estimate the body condition of humpback whales (Megaptera novaeangliae) at two feeding grounds in Canada and Norway: animal-borne tags (n = 59) and aerial photogrammetry (n = 55). Whales that had a large length-standardized projected area in overhead images (i.e. whales looked fatter) had lower estimated tissue body density (TBD) (greater lipid stores) from tag data. Linking both measurements in a Bayesian hierarchical model to estimate the true underlying (hidden) tissue body density (uTBD), we found uTBD was lower (−3.5 kg m⁻³) in pregnant females compared to adult males and resting females, while in lactating females it was higher (+6.0 kg m⁻³). Whales were more negatively buoyant (+5.0 kg m⁻³) in Norway than Canada during the early feeding season, possibly owing to a longer migration from breeding areas. While uTBD decreased over the feeding season across life-history traits, whale tissues remained negatively buoyant (1035.3 ± 3.8 kg m⁻³) in the late feeding season. This study adds confidence to the effectiveness of these independent methods to estimate the body condition of free-ranging whales.