Rice disease has serious negative effects on crop yield, and the correct diagnosis of rice diseases is the key to avoid these effects. However, the existing disease diagnosis methods for rice are neither accurate nor efficient, and special equipment is often required. In this study, an automatic diagnosis method was developed and implemented in a smartphone app. The method was developed using deep learning based on a large dataset that contained 33,026 images of six types of rice diseases: leaf blast, false smut, neck blast, sheath blight, bacterial stripe disease, and brown spot. The core of the method was the Ensemble Model in which submodels were integrated. Finally, the Ensemble Model was validated using a separate set of images. Results showed that the three best submodels were DenseNet-121, SE-ResNet-50, and ResNeSt-50, in terms of several attributes, such as, learning rate, precision, recall, and disease recognition accuracy. Therefore, these three submodels were selected and integrated in the Ensemble Model. The Ensemble Model minimized confusion among the different types of disease, reducing misdiagnosis of the disease. Using the Ensemble Model to diagnose six types of rice diseases, an overall accuracy of 91% was achieved, which is considered to be reasonably good, considering the appearance similarities in some types of rice disease. The smartphone app allowed the client to use the Ensemble Model on the web server through a network, which was convenient and efficient for the field diagnosis of rice leaf blast, false smut, neck blast, sheath blight, bacterial stripe disease, and brown spot.

To reduce the cost of production and the pollution of the environment that is due to the overapplication of herbicide in paddy fields, the location information of rice seedlings and weeds must be detected in site-specific weed management (SSWM). With the development of deep learning, a semantic segmentation method with the SegNet that is based on fully convolutional network (FCN) was proposed. In this paper, RGB color images of seedling rice were captured in paddy field, and ground truth (GT) images were obtained by manually labeled the pixels in the RGB images with three separate categories, namely, rice seedlings, background, and weeds. The class weight coefficients were calculated to solve the problem of the unbalance of the number of the classification category. GT images and RGB images were used for data training and data testing. Eighty percent of the samples were randomly selected as the training dataset and 20% of samples were used as the test dataset. The proposed method was compared with a classical semantic segmentation model, namely, FCN, and U-Net models. The average accuracy rate of the SegNet method was 92.7%, whereas the average accuracy rates of the FCN and U-Net methods were 89.5% and 70.8%, respectively. The proposed SegNet method realized higher classification accuracy and could effectively classify the pixels of rice seedlings, background, and weeds in the paddy field images and acquire the positions of their regions.

Recent advancements in deep learning techniques have pushed forward the frontiers of real photograph denoising. However, due to the inherent pooling operations in the spatial domain, current CNN-based denoisers are biased towards focusing on low-frequency representations, while discarding the high-frequency components. This will induce a problem for suboptimal visual quality as the image denoising tasks target completely eliminating the complex noises and recovering all fine-scale and salient information. In this work, we tackle this challenge from the frequency perspective and present a new solution pipeline, coined as frequency attention denoising network (FADNet). Our key idea is to build a learning-based frequency attention framework, where the feature correlations on a broader frequency spectrum can be fully characterized, thus enhancing the representational power of the network across multiple frequency channels. Based on this, we design a cascade of adaptive instance residual modules (AIRMs). In each AIRM, we first transform the spatial-domain features into the frequency space. Then, a learning-based frequency attention framework is devised to explore the feature inter-dependencies converted in the frequency domain. Besides this, we introduce an adaptive layer by leveraging the guidance of the estimated noise map and intermediate features to meet the challenges of model generalization in the noise discrepancy. The effectiveness of our method is demonstrated on several real camera benchmark datasets, with superior denoising performance, generalization capability, and efficiency versus the state-of-the-art.