Automatic and accurate estimation of disease severity is essential for food security, disease management, and yield loss prediction. Deep learning, the latest breakthrough in computer vision, is promising for fine-grained disease severity classification, as the method avoids the labor-intensive feature engineering and threshold-based segmentation. Using the apple black rot images in the PlantVillage dataset, which are further annotated by botanists with four severity stages as ground truth, a series of deep convolutional neural networks are trained to diagnose the severity of the disease. The performances of shallow networks trained from scratch and deep models fine-tuned by transfer learning are evaluated systemically in this paper. The best model is the deep VGG16 model trained with transfer learning, which yields an overall accuracy of 90.4% on the hold-out test set. The proposed deep learning model may have great potential in disease control for modern agriculture.

Plant image identification has become an interdisciplinary focus in both botanical taxonomy and computer vision. The first plant image dataset collected by mobile phone in natural scene is presented, which contains 10,000 images of 100 ornamental plant species in Beijing Forestry University campus. A 26-layer deep learning model consisting of 8 residual building blocks is designed for large-scale plant classification in natural environment. The proposed model achieves a recognition rate of 91.78% on the BJFU100 dataset, demonstrating that deep learning is a promising technology for smart forestry.

Flower grading is a significant task because it is extremely convenient for managing the flowers in greenhouse and market. With the development of computer vision, flower grading has become an interdisciplinary focus in both botany and computer vision. A new dataset named BjfuGloxinia contains three quality grades; each grade consists of 107 samples and 321 images. A multi-input convolutional neural network is designed for large scale flower grading. Multi-input CNN achieves a satisfactory accuracy of 89.6% on the BjfuGloxinia after data augmentation. Compared with a single-input CNN, the accuracy of multi-input CNN is increased by 5% on average, demonstrating that multi-input convolutional neural network is a promising model for flower grading. Although data augmentation contributes to the model, the accuracy is still limited by lack of samples diversity. Majority of misclassification is derived from the medium class. The image processing based bud detection is useful for reducing the misclassification, increasing the accuracy of flower grading to approximately 93.9%.

To compensate for the lack of charging equipment, the charging reservation mechanism has been employed in the vehicle-to-grid (V2G) system, which is vulnerable to some potential risks, such as privacy disclosure, identity impersonation, hardware modification, and so on. In this paper, we first introduce a 5G-V2G system that integrates the 5G network and the power grid to enable the remote transmission of reservation information. On this basis, we propose a scalable reservation authentication and key agreement protocol that can support access of flexible number and guarantee the confidentiality and privacy of critical reservation data. The proposed protocol uses lightweight algorithms like reverse fuzzy extraction to offload the computation overhead on the EVs and embeds physical unclonable function (PUF) in the outfield equipment to resist physical attacks. We evaluate the security properties of the proposed protocol by the formal analysis tool AVISPA and compare it with the existing schemes in terms of computing overhead, transmission overhead, signaling overhead, and security. The comprehensive results indicate that our protocol performs better in all aspects and has higher reliability and scalability. In practice, our protocol can assist the 5G charging reservation mechanism to complete efficient, lightweight, and privacy-preserving reservation information collection and authentication.

Highly efficient second harmonic generation (SHG) bridging the mid-infrared (IR) and near-IR wavelengths in a coupled hyperbolic metamaterial waveguide with a nonlinear-polymer-filled nanoscale slot is theoretically investigated. By engineering the geometrical parameters, the collinear phase matching condition is satisfied between the even hybrid modes at the fundamental frequency (3,100 nm) and the second harmonic (1,550 nm). Two modes manifest the great field overlap and the significant field enhancement in the nonlinear integration area (i.e. the slot), which leads to extreme large nonlinear coupling coefficient. For a low pumping power of 100 mW, the device length is as short as 2.19 µm and the normalized conversion efficiency comes up to more than 6.37 × 10⁵ W⁻¹cm⁻² which outperforms that of the plasmonic-based structures. Moreover, the efficient SHG can be achieved with great phase matching tolerance, i.e., a small theoretical fabrication-error sensitivity to filling ratio and a broad pump bandwidth in a compact device length of 2.19 µm using 100 mW pump. The proposed scheme links the mature near-IR devices to the mid-IR regime and have a great potential for integrated chip-scale all-optical signal processes.