The success of the ClassSR has led to a strategy of decomposing images being used for large image SR. The decomposed image patches have different recovery difficulties. Therefore, in ClassSR, image patches are reconstructed by different networks to greatly reduce the computational cost. However, in ClassSR, the training of multiple sub-networks inevitably increases the training difficulty. Furthermore, decomposing images with overlapping not only increases the computational cost but also inevitably produces artifacts. To address these challenges, we propose an end-to-end general framework, named patches separation and artifacts removal SR (PSAR-SR). In PSAR-SR, we propose an image information complexity module (IICM) to efficiently determine the difficulty of recovering image patches. Then, we propose a patches classification and separation module (PCSM), which can dynamically select an appropriate SR path for image patches of different recovery difficulties. Moreover, we propose a multi-attention artifacts removal module (MARM) in the network backend, which can not only greatly reduce the computational cost but also solve the artifacts problem well under the overlapping-free decomposition. Further, we propose two loss functions - threshold penalty loss (TP-Loss) and artifacts removal loss (AR-Loss). TP-Loss can better select appropriate SR paths for image patches. AR-Loss can effectively guarantee the reconstruction quality between image patches. Experiments show that compared to the leading methods, PSAR-SR well eliminates artifacts under the overlapping-free decomposition and achieves superior performance on existing methods (e.g., FSRCNN, CARN, SRResNet, RCAN and CAMixerSR). Moreover, PSAR-SR saves 53%-65% FLOPs in computational cost far beyond the leading methods. The code will be made available: https://github.com/dywang95/PSAR-SR.

Reference-based super-resolution (RefSR), which aims to introduce an additional high-resolution (HR) reference (Ref) image to improve the reconstruction performance of low-resolution (LR) image, has achieved great success. Existing RefSR methods rely on the texture information of the reference image to compensate for the missing information. However, the differences of scale and orientation are unavoidable when obtaining useful information from the Ref image. In addition, it is difficult to achieve a good match due to the ill-posed between the LR image and Ref image. To address these challenges, we propose a new matching module, named progressive spatial adaptation module (PSAM). PSAM is a progressive alignment model to effectively overcome the ill-pose between the LR image and the Ref image. Further, we propose a spatial correction module (SCM) to correct for scale and orientation. Meanwhile, we introduce a gradient map to further correct the matched features. In addition, we propose a new loss function MC-Loss to ensure the success of correction. Experiments show that the matching method using PSAM to directly replace the existing RefSR is significantly better than the original matching method in terms of both quantitative and qualitative results.

Increasingly serious health problems have made the usage of computed tomography surge. Therefore, algorithms for processing CT images are becoming more and more abundant. These algorithms can lessen the harm of cumulative radiation in CT technology for the patient while eliminating the noise of image caused by dose reduction. However, the mainstream CNN-based algorithms are inefficient when dealing with features in broad regions. Inspired by the large receptive field of transformer framework, this paper designs an end-to-end low-dose CT (LDCT) denoising network based on the transformer. The overall network contains a main branch and dual side branches. Specifically, the overlapping-free window-based self-attention transformer block is adopted on the main branch to realize image denoising. On the dual side branches, we propose double enhancement module to enrich edge, texture, and context information of LDCT images. Meanwhile, the receptive field of network is further enlarged after processing, which is helpful for building model's long-range dependencies. The outputs of the side branches are concatenated for enhancing information and generating high-quality CT images. In addition, to better train the network, we introduce a compound loss function including mean squared error (MSE), multi-scale perceptual (MSP), and Sobel-L1 (SL) to make the denoised image closer to the targeted norm-dose CT (NDCT) image. Lastly, we conducted experiments on two clinical datasets including abdomen, head, and chest LDCT images with 25%, 25%, and 10% of the full dose, respectively. The experimental results demonstrated that the proposed DEformer achieved better denoising performance than the existing algorithms.

Object segmentation in deep learning has been recently used for the detection of Fusarium head blight (FHB), a worldwide disease in wheat. Such method, however, cannot detect the disease with high accuracy and is difficult to be used in labelling annotation. However, object detection network can solve the above problem. The object detection network has high detection accuracy and easy for labeling. Yolov5 is an advanced object detection network, but it can’t detect the neighboring wheat ears well. So in this study, a novel method was developed based on object detection network, feature extraction and classifier to overcome these disadvantages. We combined Yolov5 object detection network with distance intersection over union non maximum suppression (DIOU-NMS) to form an improved Yolov5 object detection network. The improved YoloV5 object detection network was employed to detect and record wheat ears in images collected from field plots at two locations over 2 years. Pre-segmentation was conducted for single individual wheat ear images using threshold segmentation; HSV and CMYK color spaces were used as the baseline in each wheat ear image for extracting comprehensive color feature (CCF). The Res-Net network was used for extracting each wheat ear’s high dimension feature (HDF). CCF and HDF were then merged as the comprehensive feature (CF) of each single wheat image. The random forest was used to classify wheat ear images into healthy wheat ears and diseased wheat ears by CF and then calculate the ratio of diseased wheat ears to total wheat ears as the level of damage caused by FHB. The results of performance evaluation of the proposed method in two different locations and years demonstrate its strong robustness in both time and spatial domains to effectively detect the levels of damage caused by FHB under the complex field background conditions. The average detection accuracy and detection time were 90.67% and 0.73 ms, respectively. The average accuracies of counting total wheat ears and diseased wheat ears were 96.16% and 81.66%, respectively. The improved YoloV5 method developed from this study can be used as a quick, efficient, and convenient tool for assessment of the levels of damage caused by FHB in wheat under field conditions.

针对传统农作物病害识别方法依靠人工提取特征,步骤复杂且低效,难以实现在田间环境下识别的问题,提出一种多尺度卷积结构与注意力机制结合的农作物病害识别模型.该研究在残差网络(ResNet 18)的基础上进行改进,引入Inception模块,利用其多尺度卷积核结构对不同尺度的病害特征进行提取,提高了特征的丰富度.在残差结构的基础上加入注意力机制SE-Net(Squeeze-and-excitation networks),增强了有用特征的权重,减弱了噪声等无用特征的影响,进一步提高特征提取能力并且增强了模型的鲁棒性.实验结果表明,改进后的多尺度注意力残差网络模型(Multi-Scale-SE-ResNet18)在复杂田间环境收集的8种农作物病害数据集上的平均识别准确率达到95.62％,相较于原ResNet18模型准确率提高10.92个百分点,模型占用内存容量仅为44.2 MB.改进后的Multi-Scale-SE-ResNet18具有更好的特征提取能力,可以提取到更多的病害特征信息,并且较好地平衡了模型的识别精度与模型复杂度,可为田间环境下农作物病害识别提供参考.

Fast and accurate counting of wheat ears in field conditions is a key element for determining wheat yield. To obtain the number of wheat ears in a field, we propose a new counting algorithm based on computer vision. This algorithm counts wheat ears in remote images through semantic segmentation regression network (SSRNet). SSRNet is a multistage convolutional neural network that we propose to achieve counting problems through regression. In SSRNet, first, the original image is cropped to increase the amount of data. This method effectively solves the small sample dataset. Next, based on the cropping results, we build a fully convolutional neural network (FCNN) to segment wheat ears in field conditions. FCNN increases the accuracy of wheat ears counting by accurately segmenting wheat ears in a complex background. Then, we build a regression convolutional neural network (RCNN) to count wheat ears based on the segmentation results of FCNN. In RCNN, we propose a new activation function positive rectification linear unit (PrLU) to process the last layer of the fully connected layer, so that RCNN can effectively count the number of wheat ears in the image. Finally, a counting strategy is proposed to count the number of wheat ears in the original image. To verify the counting performance of SSRNet, we compare the counting result of SSRNet with the real value of manual statistics. The results show that the average accuracy (Acc), R-2, and root mean squared error (RMSE) of the SSRNet count results on the test set in this article are 0.980, 0.996, and 9.437, respectively. It can be seen from the results that our proposed method can accurately count wheat ears in field conditions. At the same time, the counting time (0.11 s) shows that SSRNet can quickly estimate the number of wheat ears in field conditions. We concluded that this study can provide important technical support for the high-throughput field wheat ears counting task in large-scale phenotyping work.

A fast and nondestructive method for recognizing the severity of wheat Fusarium head blight (FHB) can effectively reduce fungicide use and associated costs in wheat production. This study proposed a feature fusion method based on deep convolution and shallow features derived from the high-resolution digital Red-green-blue (RGB) images of wheat FHB at different disease severity levels. To test the robustness of the proposed method, the RGB images were taken under different influence factors including light condition, camera shooting angle, image resolution, and crop growth period. All images were preprocessed to eliminate background noises to improve recognition accuracy. The AlexNet model parameters trained by the ImageNet 2012 dataset were transferred to the test dataset to extract the deep convolution feature of wheat FHB. Next, the color and texture features of wheat ears were extracted as shallow features. Then, the Relief-F algorithm was used to fuse the deep convolution feature and shallow features as the final FHB features. Finally, the random forest was used to classify and identify the features of different FHB severity levels. Results show that the recognition accuracy of the proposed fusion feature model was higher than those of models using other features in all conditions. The highest recognition accuracy of severity levels was obtained when images were taken under indoor conditions, with high resolution (12 MB pixels), at 90° shooting angle during the crop filling period. The Relief-F algorithm assigned different weights to the features under different influence factors; it made the fused feature model more robust and improved the ability to recognize wheat FHB severity levels using RGB images.

为了利用数码影像快速提取马铃薯覆盖度,首先,利用植被覆盖度提取算法从地面数码影像中获取马铃薯覆盖度实测值;然后,通过植被指数提取法和最大似然监督分类法对无人机数码影像进行处理,分别获取各个研究小区的马铃薯覆盖度;提出利用颜色转换空间HSI(H-A法)从无人机数码影像中快速提取马铃薯覆盖度;最后,对H-A法、最大似然监督分类法和植被指数提取法3种方法的计算结果进行精度比较.结果 表明,H-A法估测的植被覆盖度的精度最高,均达到0.91以上,拟合函数拟合度为0.97;最大似然监督分类法次之,最低精度为0.75,拟合度为0.82;植被指数提取法最差,最低精度为0.74,拟合度为0.74.

Corn is one of the most important food crops in the world, but there are many kinds of corn diseases, and it is difficult to diagnose by planting personnel based on experience. However, the misdiagnosis reduces production efficiency. With the development of computer technology, the use of deep learning and image recognition technology for plant disease detection has become an important research direction. We propose a convolutional neural network based on data augmentation combined with transfer learning to identify corn leaf disease models. The algorithm first increases the data by means of data augmentation to improve the generalization and accuracy of the model, and builds a convolutional neural network model based on transfer learning. Then, it uses the model for training, accelerates the training process of the convolutional neural network, and uses test dataset feedback network training results. In this study, the corn leaf images in PlantVillage were used as the dataset of our experiment to classify the four categories which consists of Corn Gray leaf spot, Corn Common rust, Corn Northern Leaf Blight and healthy leaves. We first obtained our optimization model by fine-tuning the GoogLeNet pre-training network, adjusting parameters such as optimizer and learning rate. Then we trained the optimization model, like the original GoogLeNet network, ResNet18, Vgg16, and Vgg19 networks based on transfer learning and compare the results. The results show that by using our optimized model, the average recognition accuracy of corn disease which consists of corn common leaf rust, Corn Common rust, Corn Northern Leaf Blight and healthy leaves reached 97.6%, and the recognition accuracy of each category was greater than 95%. Compared with the original GoogLeNet model, the highest accuracy rate is improved by 5.9%. Also, the effect is better when compared with other networks based on transfer learning. Our model provides new ideas for the identification of diseases, insect pests of corn and other crops.

Accurate counting of wheat ears in field conditions is vital to predict yield and for crop breeding. To quickly and accurately obtain the number of wheat ears in a field, we propose herein a method to count wheat ears based on fully convolutional network (FCN) and Harris corner detection. The technical procedure consists essentially of 1) constructing a dataset of wheat-ear images from acquired red-green-blue (RGB) images; 2) training a FCN as the wheat-ear segmentation model by using the constructed image dataset; 3) preparing testing images and inputting them into the segmentation model to get the initial segmentation results; 4) binarizing the initial segmentation by using the Otsu algorithm (to facilitate subsequent processing); and 5) applying Harris corner detection after extracting the wheat-ear skeleton to obtain the number of wheat ears in the images. The segmentation results show that the proposed FCN-based segmentation model segments wheat ears with an average accuracy of 0.984 and at low computational cost. An average of only 0.033 s is required to segment a $256\times 256$ -pixel wheat-ear image. Moreover, the segmentation result is improved by nearly 10% compared with the previous segmentation methods under conditions of wheat-ear occlusion, leaf occlusion, uneven illumination, and soil disturbance. Subsequently, the proposed counting method achieves good results, with an average accuracy of 0.974, a coefficient of determination (R 2 ) of 0.983, and a root mean square error (RMSE) of 14.043. These metrics are all improved by 10% compared with the previous methods. These results show that the proposed method accurately counts wheat ears even under conditions of wheat-ear adhesion. Furthermore, the results provide an important technique for studying wheat phenotyping.

Fusarium head blight (FHB), one of the most important diseases of wheat, mainly occurs in the ear. Given that the severity of the disease cannot be accurately identified, the cost of pesticide application increases every year, and the agricultural ecological environment is also polluted. In this study, a neural network (NN) method was proposed based on the red-green-blue (RGB) image to segment wheat ear and disease spot in the field environment, and then to determine the disease grade. Firstly, a segmentation dataset of single wheat ear was constructed to provide a benchmark for the segmentation of the wheat ear. Secondly, a segmentation model of single wheat ear based on the fully convolutional network (FCN) was established to effectively realize the segmentation of the wheat ear in the field environment. An FHB segmentation algorithm was proposed based on a pulse-coupled neural network (PCNN) with K-means clustering of the improved artificial bee colony (IABC) to segment the diseased spot of wheat ear by automatic optimization of PCNN parameters. Finally, the disease grade was calculated using the ratio of the disease spot to the whole wheat ear. The experimental results show that: (1) the accuracy of the segmentation model for single wheat ear constructed in this study is 0.981. The segmentation time is less than 1 s, indicating that the model can quickly and accurately segment wheat ear in the field environment; (2) the segmentation method of the disease spot performed under each evaluation indicator is improved compared with the traditional segmentation methods, and the accuracy is 0.925 in the disease severity identification. These research results can provide important reference value for grading wheat FHB in the field environment, which also can be beneficial for real-time monitoring of other crops’ diseases under near-Earth remote sensing.

Wheat (Triticum aestivum L.) is one of the three major cereals worldwide. The FusaHum graminearum Sehw., special fugus always damages the wheat ear, and produces vomitoxin,is difficult to control and prevent, and seriously threatens the health of humans, animals and China's food security. Currently, rapidly, accurately and non-destructively diagnostic devices or systems for this disease have not been disclosed. In this study, the infected ears with different severities were picked up in key growth stages. The diseased area of wheat ear was extracted using hypergreen characteristic, and a total of 30 features of infected ears were chosen including color (Lab, HSI, HSV, YCbCr color space), texture (LBP and LLE dimension reduction), and shape (squareness, shape complexity, and eccentricity). Then using the competitive adaptive re-weighted sampling (CARS) and rough set algorithm (RS) to screen the characteristics of the diseased ear, the four characteristics with the largest contribution were determined to establish the CARS-SVM and CARS-RS-SVM models respectively. The study found that the recognition rate of CARS-SVM model is 85.4%, while CARS-RS-SVM model is 92.7%. Thus the CARS-RS-SVM was thought of as the optimal model by two indicators of identification accuracy. On the basis, a wheat scab diagnosis system based on Android mobile phone was constructed. It consists of three parts - Clients, Service-Terminal and Database. The Client was designed by Android Studio and its functions mainly include image acquisition, image storage, GPS positioning, image uploading and diagnostic results display. The Service-Terminal was completed by the mixed programming of Myeclipse and Matlab software, and Tomcat was used as the Server. It mainly implements the functions of image receiving, image preprocessing, feature extraction and selection, and classifier modeling. The MySQL was used to establish two databases: the “Disease Characteristics Database” and the “Disease Diagnosis Knowledge Base”. Finally, through samples testing and validating, the Android-based mobile terminal can real-time collect the image of Fusarium head blight and upload the server. After the target image was processed and compared by the “Disease Characteristics Database”, the appropriate diagnostic knowledge was selected from the “Disease Diagnosis Knowledge Base” and feedbacked to the client. In summary, the results of this study showed that it was helpful for the rapid and non-destructive investigation of infected FHB in the field, and it would provide a reference for the study of other crop diseases, facilitate the application and development of new technologies such as artificial intelligence and big data in agriculture.

Powdery mildew is one of the most widely destructive plant diseases, particularly infecting winter wheat. Early detection of wheat powdery mildew is of importance, which is useful to reduce economic costs and environmental pollution. However, difficulties emerge at early development stages, due to slight variations of the characteristic symptoms. Fortunately, hyperspectral reflectance imaging has been proven as a powerful tool to detecting early disease severity in plant. In this study, hyperspectral imagery data of leaves were acquired at early stages of the disease in winter wheat. It was demonstrated that early powdery mildew could induce observable spectral changes in both visible and near infrared regions. Given that, powdery mildew indices (PMI) were constructed and showed the capability of distinguishing between normal and diseased leaves, although it displayed poor effects for differentiating disease-damaged levels of early powdery mildew and estimating disease severity. However, further study was carried out by combination of hyperspectral vegetation indices closely related to plant diseases. It was noticeable that the model of three indices of PRI, PSRI and ARI significantly increased the classification accuracy of various early disease levels, and the regression model of PMI, PSRI and ARI apparently improved the estimation accuracy of disease severity. These valuable results could be used to prevent the development and the spread of the disease, and particularly beneficial to develop a portable or automated sensor in precision agriculture. (C) 2018 Friends Science Publishers