Virtual Issue

智慧植保合辑Smart Plant Protection

Default Latest Most Read

Please wait a minute... For Selected: Download Citations EndNote Ris BibTeX Toggle Thumbnails Select Automatic image segmentation method for cotton leaves with disease under natural environment ZHANG Jian-hua, KONG Fan-tao, WU Jian-zhai, HAN Shu-qing, ZHAI Zhi-fen 2018, 17 (08): 1800-1814.   DOI: 10.1016/S2095-3119(18)61915-X Abstract （325）      PDF （31718KB）（105）       In order to improve the image segmentation performance of cotton leaves in natural environment, an automatic segmentation model of diseased leaf with active gradient and local information is proposed.  Firstly, a segmented monotone decreasing edge composite function is proposed to accelerate the evolution of the level set curve in the gradient smooth region.  Secondly, canny edge detection operator gradient is introduced into the model as the global information.  In the process of the evolution of the level set function, the guidance information of the energy function is used to guide the curve evolution according to the local information of the image, and the smooth contour curve is obtained.  And the main direction of the evolution of the level set curve is controlled according to the global gradient information, which effectively overcomes the local minima in the process of the evolution of the level set function.  Finally, the Heaviside function is introduced into the energy function to smooth the contours of the motion and to increase the penalty function ?(x) to calibrate the deviation of the level set function so that the level set is smooth and closed.  The results showed that the model of cotton leaf edge profile curve could be obtained in the model of cotton leaf covered by bare soil, straw mulching and plastic film mulching, and the ideal edge of the ROI could be realized when the light was not uniform.  In the complex background, the model can segment the leaves of the cotton with uneven illumination, shadow and weed background, and it is better to realize the ideal extraction of the edge of the blade.  Compared with the Geodesic Active Contour (GAC) algorithm, Chan-Vese (C-V) algorithm and Local Binary Fitting (LBF) algorithm, it is found that the model has the advantages of segmentation accuracy and running time when processing seven kinds of cotton disease leaves images, including uneven lighting, leaf disease spot blur, adhesive diseased leaf, shadow, complex background, unclear diseased leaf edges, and staggered condition.  This model can not only conduct image segmentation of cotton leaves under natural conditions, but also provide technical support for the accurate identification and diagnosis of cotton diseases. Reference | Related Articles | Metrics Select MmNet: Identifying Mikania micrantha Kunth in the wild via a deep Convolutional Neural Network QIAO Xi, LI Yan-zhou, SU Guang-yuan, TIAN Hong-kun, ZHANG Shuo, SUN Zhong-yu, YANG Long, WAN Fang-hao, QIAN Wan-qiang 2020, 19 (5): 1292-1300.   DOI: 10.1016/S2095-3119(19)62829-7 Abstract （123）      PDF in ScienceDirect       Mikania micrantha Kunth is an invasive alien weed and known as a plant killer around the world.  Accurately and rapidly identifying M. micrantha in the wild is important for monitoring its growth status, as this helps management officials to take the necessary steps to devise a comprehensive strategy to control the invasive weed in the identified area.  However, this approach still mainly depends on satellite remote sensing and manual inspection.  The cost is high and the accuracy rate and efficiency are low.  We acquired color images of the monitoring area in the wild environment using an Unmanned Aerial Vehicle (UAV) and proposed a novel network -MmNet- based on a deep Convolutional Neural Network (CNN) to identify M. micrantha in the images.  The network consists of AlexNet Local Response Normalization (LRN), along with the GoogLeNet and continuous convolution of VGG inception models.  After training and testing, the identification of 400 testing samples by MmNet is very good, with accuracy of 94.50% and time cost of 10.369 s.  Moreover, in quantitative comparative analysis, the proposed MmNet not only has high accuracy and efficiency but also simple construction and outstanding repeatability.  Compared with recently popular CNNs, MmNet is more suitable for the identification of M. micrantha in the wild.  However, to meet the challenge of wild environments, more M. micrantha images need to be acquired for MmNet training.  In addition, the classification labels need to be sorted in more detail.  Altogether, this research provides some theoretical and scientific basis for the development of intelligent monitoring and early warning systems for M. micrantha and other invasive species.  Reference | Related Articles | Metrics Select Using an image segmentation and support vector machine method for identifying two locust species and instars Shuhan LU, YE Si-jing 2020, 19 (5): 1301-1313.   DOI: 10.1016/S2095-3119(19)62865-0 Abstract （59）      PDF in ScienceDirect       Locusts are agricultural pests around the world.  To cognize how locust distribution density and community structure are related to the hydrothermal and vegetation growth conditions of their habitats and thereby providing rapid and accurate warning of locust invasions, it is important to develop efficient and accurate techniques for acquiring locust information.  In this paper, by analyzing the differences between the morphological features of Locusta migratoria manilensis and Oedaleus decorus asiaticus, we proposed a semi-automatic locust species and instar information detection model based on locust image segmentation, locust feature variable extraction and support vector machine (SVM) classification.  And we subsequently examined its applicability and accuracy based on sample image data acquired in the field.  Locust image segmentation experiment showed that the proposed GrabCut-based interactive segmentation method can be used to rapidly extract images of various locust body parts and exhibits excellent operability.  In a locust feature variable extraction experiment, the textural, color and morphological features of various locust body parts were calculated.  Based on the results, eight feature variables were selected to identify locust species and instars using outlier detection, variable function calculation and principal component analysis.  An SVM-based locust classification experiment achieved a semi-automatic detection accuracy of 96.16% when a polynomial kernel function with a penalty factor parameter c of 2 040 and a gamma parameter g of 0.5 was used.  The proposed detection model exhibits advantages such as high applicability and accuracy when it is used to identify locust instars of L. migratoria manilensis and O. decorus asiaticus, and it can also be used to identify other species of locusts. Reference | Related Articles | Metrics Select Predicting the potential geographic distribution of Bactrocera bryoniae and Bactrocera neohumeralis (Diptera: Tephritidae) in China using MaxEnt ecological niche modeling Jing Wan, QI Guo-jun, MA Jun, Yonglin Ren, WANG Rui, Simon MCKIRDY 2020, 19 (8): 2072-2082.   DOI: 10.1016/S2095-3119(19)62840-6 Abstract （107）      PDF in ScienceDirect       Bactrocera bryoniae and Bactrocera neohumeralis are highly destructive and major biosecurity/quarantine pests of fruit and vegetable in the tropical and subtropical regions in the South Pacific and Australia.  Although these pests have not established in China, precautions must be taken due to their highly destructive nature.  Thus, we predicted the potential geographic distribution of B.?bryoniae and B. neohumeralis across the world and in particular China by ecological niche modeling of the Maximum Entropy (MaxEnt) model with the occurrence records of these two species. Bactrocera bryoniae and B. neohumeralis exhibit similar potential geographic distribution ranges across the world and in China, and each species was predicted to be able to distribute to over 20% of the globe.  Globally, the potential geographic distribution ranges for these two fruit fly species included southern Asia, the central and the southeast coast of Africa, southern North America, northern and central South America, and Australia.  While within China, most of the southern Yangtze River area was found suitable for these species.  Notably, southern China was considered to have the highest risk of B. bryoniae and B. neohumeralis invasions.  Our study identifies the regions at high risk for potential establishment of B. bryoniae and B. neohumeralis in the world and in particular China, and informs the development of inspection and biosecurity/quarantine measures to prevent and control their invasions.   Reference | Related Articles | Metrics Select Development of an automatic monitoring system for rice light-trap pests based on machine vision YAO Qing, FENG Jin, TANG Jian, XU Wei-gen, ZHU Xu-hua, YANG Bao-jun, Lü Jun, XIE Yi-ze, YAO Bo, WU Shu-zhen, KUAI Nai-yang, WANG Li-jun 2020, 19 (10): 2500-2513.   DOI: 10.1016/S2095-3119(20)63168-9 Abstract （62）      PDF in ScienceDirect       Monitoring pest populations in paddy fields is important to effectively implement integrated pest management.  Light traps are widely used to monitor field pests all over the world.  Most conventional light traps still involve manual identification of target pests from lots of trapped insects, which is time-consuming, labor-intensive and error-prone, especially in pest peak periods.  In this paper, we developed an automatic monitoring system for rice light-trap pests based on machine vision.  This system is composed of an intelligent light trap, a computer or mobile phone client platform and a cloud server.  The light trap firstly traps, kills and disperses insects, then collects images of trapped insects and sends each image to the cloud server.  Five target pests in images are automatically identified and counted by pest identification models loaded in the server.  To avoid light-trap insects piling up, a vibration plate and a moving rotation conveyor belt are adopted to disperse these trapped insects.  There was a close correlation (r=0.92) between our automatic and manual identification methods based on the daily pest number of one-year images from one light trap.  Field experiments demonstrated the effectiveness and accuracy of our automatic light trap monitoring system. Reference | Related Articles | Metrics Select A rapid, low-cost deep learning system to classify strawberry disease based on cloud service YANG Guo-feng, YANG Yong, HE Zi-kang, ZHANG Xin-yu, HE Yong 2022, 21 (2): 460-473.   DOI: 10.1016/S2095-3119(21)63604-3 Abstract （115）      PDF in ScienceDirect       Accurate and timely classification of diseases during strawberry planting can help growers deal with them in timely manner, thereby reducing losses.  However, the classification of strawberry diseases in real planting environments is facing severe challenges, including complex planting environments, multiple disease categories with small differences, and so on.  Although recent mobile vision technology based deep learning has achieved some success in overcoming the above problems, a key problem is how to construct a non-destructive, fast and convenient method to improve the efficiency of strawberry disease identification for the multi-region, multi-space and multi-time classification requirements.  We develop and evaluate a rapid, low-cost system for classifying diseases in strawberry cultivation.  This involves designing an easy-to-use cloud-based strawberry disease identification system, combined with our novel self-supervised multi-network fusion classification model, which consists of a Location network, a Feedback network and a Classification network to identify the categories of common strawberry diseases.  With the help of a novel self-supervision mechanism, the model can effectively identify diseased regions of strawberry disease images without the need for annotations such as bounding boxes.  Using accuracy, precision, recall and F1 to evaluate the classification effect, the results of the test set are 92.48, 90.68, 86.32 and 88.45%, respectively.  Compared with popular Convolutional Neural Networks (CNN) and five other methods, our network achieves better disease classification effect.  Currently, the client (mini program) has been released on the WeChat platform.  The mini program has perfect classification effect in the actual test, which verifies the feasibility and effectiveness of the system, and can provide a reference for the intelligent research and application of strawberry disease identification. Reference | Related Articles | Metrics Select Intelligent diagnosis of northern corn leaf blight with deep learning model PAN Shuai-qun, QIAO Jing-fen, WANG Rui, YU Hui-lin, WANG Cheng, Kerry TAYLOR, PAN Hong-yu 2022, 21 (4): 1094-1105.   DOI: 10.1016/S2095-3119(21)63707-3 Abstract （141）      PDF in ScienceDirect       Maize (Zea mays L.), also known as corn, is the third most cultivated crop in the world.  Northern corn leaf blight (NCLB) is a globally devastating maize foliar disease caused by Setosphaeria turcica (Luttrell) Leonard and Suggs.  Early intelligent diagnosis and warning is an effective and economical strategy to control this disease.  Today, deep learning is beginning to play an essential role in agriculture.  Notably, deep convolutional neural networks (DCNN) are amongst the most successful machine learning techniques in plant disease detection and diagnosis.  Our study aims to identify NCLB in the maize-producing area in Jilin Province based on several DCNN models.  We established a database of 985 leaf images of healthy and infected maize and applied data augmentation techniques including image segmentation, image resizing, image cropping, and image transformation, to expand to 30 655 images.  Several proven convolutional neural networks, such as AlexNet, GoogleNet, VGG16, and VGG19, were then used to identify diseases.  Based on the best performance of the DCNN pre-trained model GoogleNet, some of the recent loss functions developed for deep facial recognition tasks such as ArcFace, CosFace, and A-Softmax were applied to detect NCLB.  We found that a pre-trained GoogleNet architecture with the Softmax loss function can achieve an excellent accuracy of 99.94% on NCLB diagnosis.  The analysis was implemented in Python with two deep learning frameworks, Pytorch and Keras.  The techniques, training, validation, and test results are presented in this paper.  Overall, our study explores intelligent identification technology for NCLB and effectively diagnoses NCLB from images of maize. Reference | Related Articles | Metrics