Search

Copyright of Science Technology & Engineering is the property of Chinese Society of Technology Economics and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)

Detection of crop pest has widely been one of the most challenges in modern agriculture, due to the intra- and inter-class pests in the field with various colors, sizes, shapes, postures, positions, and complex backgrounds. Convolutional Neural Network (CNN) has presented an excellent performance on the detection and recognition of complex images. However, the current CNN models cannot adapt to the geometric deformation of pests. In this study, a deformable VGG-16 (DVGG-16) model was constructed and then applied for the detection of crop pest in the field. The framework consisted of six convolutional layers, four deformable convolutional layers, five pooling layers, and one global average pooling layer. Furthermore, the network training was utilized to speed up the global average pooling operation, instead of three fully connected layers of VGG-16. Four convolutional layers in VGG-16 were replaced by four deformable convolutional layers, in order to improve the characteristic expression ability of network and the practicality of VGG-16 to insect image deformation. Moreover, a global pooling layer was used instead of three fully connected layers of VGG-16, in order to reduce the number of the training parameters, while accelerate the network training speed free of the over-fitting. The offset was added in the deformable convolution unit, thereby to serve one part of DVGG-16 structure. Among them, another parallel standard convolution unit was used to calculate and then learn end-to-end through gradient backpropagation. Subsequently, the size of deformable convolution kernels and position were adjusted, according to the current need to identify the dynamic image content of crop pests, particularly suitable for different shapes, sizes, and other geometric deformation of the object. Moreover, data augmentation was performed on the original dataset to increase the number of training samples. A series operations were also included for the better generalization ability and robustness of model, such as bilinear interpolation, cropping and rotating images, and adding salt-pepper noise to the images. A parallel convolution layer was used in DVGG-16 to learn the offset corresponding to the input feature map. The constraint was easily broken for the regular grid of normal convolution, where an offset was added at the corresponding position of each sampling point, while the arbitrary sampling was performed around the sampling location. More importantly, the deformable convolution was greatly contributed to the DVGG-16 model for better suitable for various insect images with different shapes, states, and sizes. An image database of actual field pest was evaluated to compare with two feature extraction and two deep learning, including image-based Orchard Insect Automated Identification (IIAI), Local Mean Color Feature and Support Vector Machine (LMCFSVM), Improved Convolutional Neural Network (ICNN), and VGG-16. Specifically, the detection accuracy of DVGG-16 was 91.14%, which was 28.60 and 26.97 percentage higher than that of IIAI and LMCFSVM, and 7.72 and 9.01 percentage higher than that of ICNN and VGG-16 based models, respectively. The training time of DVGG-16 was 7.98 h longer than that of the ICNN, because the deformable convolution operation was realized by bilinear interpolation, which resulted in the increase of computational complexity and training time of DVGG-16 compared with ICNN. The test time of DVGG-16 based model was 0.02 and 0.17 s faster than that ICNN and VGG-16 based models, respectively. Consequently, the DVGG-16 was effective and feasible to detect the variable pests in the field. The finding can provide a strong reference for the effective detection of pests in the complex field background, further to realize the feature extraction of irregular field insect images without changing the spatial resolution. [ABSTRACT FROM AUTHOR]

Cucumber disease leaf image segmentation is an important step in disease detection and disease type recognition. To overcome the shortcomings of the classical disease leaf segmentation methods, image semantic segmentation algorithm based on the Fully Convolution Networks (FCNs) had been widely used in the automatic segmentation of disease leaf images in the complex background. FCNs replaced the last three fully-connected layers with three convolutional layers so that the input image with any size could be accepted. FCNs classified images at the pixel level, resolving the problem of semantic segmentation at the semantic level. FCNs utilized the de-convolutional layer to upsample the feature map of the last convolutional layer and restored it to the same size of the input image so that each pixel could be generated. At the same time, the spatial information of the original input image was retained. Then, the pixel-by-pixel classification was carried out on the above feature maps. The disadvantages of FCNs were that 1) the segmented images by FCNs were still not precise enough. Although the result of 8 times sampling was much better than 32 times sampling, the result of upsampling was still blurred and smooth, and was insensitive to the details of the image; 2) Classification of each pixel did not fully consider the relationship between the pixels. The spatial regularization steps used in the usual segmentation methods based on pixel classification were neglected and lack of spatial consistency. Aiming at the low recognition accuracy problem of the traditional disease leaf image segmentation methods, the Multi-Scale Fusion Convolutional Neural Networks (MSF-CNNS) were proposed for cucumber disease leaf image segmentation. MSF-CNNs consisted of Encoder Networks (ENs) and Decoder Networks (DNs). ENs were composed of a multi-scale Convolutional Neural Networks to extract multi-scale information of images of disease leaves. DNs were a nine-point bilinear interpolation algorithm to restore the size and resolution of the input image. In the process of the model training, a transfer learning method with the gradual adjustment was used to accelerate the training speed and segmentation accuracy of the network model. The architecture of MSF-CNNs is similar to U-Net and SegNet, mainly including encoder networks and decoder networks. However, to extract the multi-scale information of the input image, a multilevel parallel structure was introduced into the encoding network, while a multi-scale connection was introduced into the decoding network. In the specific coding network, the multi-column parallel CNNs could be used to extract the multi-scale features of the image of crop disease leaves. In the decoding network, the size and resolution of the image were restored by introducing the nine-point bisector linear interpolation algorithm as the deconvolution interpolation method. In the structure of the overall network model, skip join was used to pass the characteristic information extracted from different convolutional layers, and batch normalization operation was introduced to alleviate the gradient dispersion phenomenon of the model. Segmentation experiments were carried out on the image database of cucumber disease leaves under the complex background and compared with the existing deep learning models, such as FCNs, SegNet, U-Net, and DenseNet. The results on the cucumber disease leaf image dataset validated that the proposed method met the needs of the cucumber disease leaf image segmentation in the complex environment, with pixel-classification accuracy of 92.38%, the average accuracy of 93.12%, mean intersection over the union of 91.36 and frequency weighted intersection over the union of 89.76%. Compared with FCNs, SegNet, U-NET, and DenseNet, the average accuracy of the proposed method is improved by 13.00%, 10.74%, 10.40%, 10.08%, and 6.40%, respectively. After using the progressive learning training method, the training time was reduced by 0.9 h. The results showed that the proposed method was effective for the image segmentation of the cucumber disease leaves in a complex environment, and could provide technical support for further research on cucumber disease detection and identification. [ABSTRACT FROM AUTHOR]

Copyright of Journal of Computer Engineering & Applications is the property of Beijing Journal of Computer Engineering & Applications Journal Co Ltd. and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)

Stripe rust disease of wheat has seriously threaten to the yield and quality of wheat. The disease prediction is a key step of disease control. However, the complex and changing environmental information that occurred concurrently wheat disease make disease prediction difficult. The prediction data of actual crop diseases usually come from environmental factors of field crop growth, or the database of agricultural scientific research institutions and enterprises, where the text data captured from the corpus of agricultural websites. These data cannot be effectively managed and utilized by traditional database and crop prediction methods, owing to the structured, unstructured, and semi-structured data with large amount, multi-source, heterogeneous, noise and redundancy. Knowledge Graph (KG) can be represented with big data and metadata of crop diseases, due to its powerful data fusion, organization, and management functions. Deep Learning (DL) model-based prediction method of crop disease can directly utilize structured data from traditional database systems, but cannot directly utilize large amounts of unstructured and semi-structured data. Most DL models fail to consider the context of text content and relevance between words, leading to a low ability of feature expression in extraction or learning. The memory of Long Short Term Memory (LSTM) model normally originates from the front to the back, leading to effectively capture in remembering contextual information, but not for contextual information from the back. Therefore, Bi-directional LSTM (Bi-LSTM) model has been added a reverse LSTM layer on top to transfer information from back to forward, indicating that Bi-LSTM can be used to effectively learn the characteristic representation of environmental factor data of crop diseases. However, the sensitivity of Bi-LSTM model to environmental factors is low, and thereby the predicted results can be difficult to understand. In this study, A prediction method of wheat rust disease was proposed using the combination of Bi-LSTM model and KG, in order to enhance the semantic information of original environmental factors and interpretability of Bi-LSTM. In the knowledge driven Bi-directional LSTM model, a physical link disambiguation and KG embedded extraction can be used to capture disease KG of structured disease knowledge, disease in the text, and described the condition of key vector with the corresponding knowledge entities vector as Bi-directional LSTM multichannel input, in the process of convolution and from the aspects of semantics and knowledge represent different types of diseases. First, KG of wheat disease was constructed, while the environmental information related to the occurrence of wheat disease was transformed into the feature vector. Then, a prediction model of wheat stripe rust was established based on Bi-LSTM. The proposed method was validated on the disease dataset of wheat stripe rust. The prediction accuracy of wheat stripe rust was 93.21%, indicating 4.5 percentage points higher than that of only Bi-LSTM based method of disease prediction. The proposed algorithm can well predict wheat disease, further to provide scientific basis for forecasting, warning, and comprehensive control of wheat rust disease. [ABSTRACT FROM AUTHOR]

Copyright of Journal of Agricultural Science & Technology (1008-0864) is the property of Journal of Agricultural Science & Technology and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)

Copyright of Transactions of the Chinese Society of Agricultural Engineering is the property of Chinese Society of Agricultural Engineering and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)

Extracting the optimal low-dimensionality classifying features from palmprint is always an important research area in the palmprint recognition field. On the basis of the rich texture features of palmprint image. this paper proposed a palmprint recognition method by combining weighted adaptive CS-LBP and local discriminant projection (LDP). Firstly, this method divided the region of interest (KOI) image of palmprint into blocks with uniform size.and computed the texture feature histogram and weighted value of each block by the adaptive CS-LBP (ACS-LBP) algorithm. Then it obtained the weighted texture feature histogram vector by connecting all weighted histograms seriaDy. Secondly, it reduced the feature vector by the local discriminant projection algorithm. Finally, it used the K-nearest neighborhood classifier to perform the palmprint identification. The proposed method was tested and compared with the existing algorithms on a public palmprint database. The recognition rate is over 97%. The experimental results show that the proposed method is not only effective and feasible for palmprint recognition, but also can be clearly interpreted. [ABSTRACT FROM AUTHOR]

Copyright of Transactions of the Chinese Society of Agricultural Engineering is the property of Chinese Society of Agricultural Engineering and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)

The diseases and insect pests of greenhouse winter jujube are one of the main factors that restrict the yield and quality of winter jujube. The timely prediction of the jujube diseases and insect pests is the prerequisite to prevent and control diseases and insect pests. It is difficult to establish an accurate forecasting model of diseases and insect pests using traditional mathematical method and neural network (NN) because of many complex factors that lead to the occurrence of diseases and insect pests of winter jujube, including the meteorological conditions (such as temperature, sunlight, humidity), soil conditions (such as moisture, soil heavy metals), and biological characteristics (such as roots, leaves). During the process of forecasting model training, due to the defects of artificial design features and the unpredictable complexity in the design process, the accuracy of disease and insect pest prediction and the efficiency of the design features can't have a lot of space of ascension. It is possible to automatically forecast diseases and insect pests of winter jujube with the development of agricultural IOT (Internet of Things), smart camera equipment, high performance and large capacity data storage, computer and network technology as well as the massive complex data processing technology. Faced with the problem of complexity and uncertainty of diseases and insect pests prediction of winter jujube, a forecasting model of winter jujube diseases and insect pests is proposed based on the modified deep belief network (DBN). Due to the merits of the DBN, the prediction model of disease and insect pest based on modified DBN can not only utilize 20 kinds of environmental information data, but also introduce the similarity between the prior information and the constraints of the current information. The modified DBN consists of a visible input layer, several hidden layers, and an output layer. The visible layer inputs the data, whose range has been normalized into [0,1]; the hidden layers are invisible, in which binary values are used, and activated by the sigmoid kernel function. Via simulating neural connecting structure of human brain and introducing the supervised information by restricting the similarity between feature vectors in the learning process, the proposed model can automatically learn senior nonlinear hierarchical combination features from the environmental information of winter jujube growth, which is suitable for data classification and importing high-level features into traditional BP (back propagation) neural network classifier to improve the disease forecasting precision. The disease and insect pest prediction is conducted by BP network in the top level of DBN. Experiments on the actual database of disease and insect pest of greenhouse winter jujube are performed. After a large number of training samples and training times, the prediction accuracy rate of diseases and insect pests is greatly improved. The accuracy rate of forecasting result is over than 84%. The experimental results show that the proposed model has provided a technical basis and support for the automatic crop disease forecasting with environmental information obtained in fields, and has great application prospect in disease and insect pest prediction of greenhouse winter jujube. As there are many factors affecting crop diseases, practically, some factors vary with the time, how to use the environmental information of crop growth to build a powerful and practical crop disease forecasting method still needs further study. [ABSTRACT FROM AUTHOR]

Copyright of Petroleum Geology & Recovery Efficiency is the property of Petroleum Geology & Recovery Efficiency and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)