There is a huge gap between the output and demand of soybean in China. How to improve the seed vigor of soybean has always been a research focus. Low temperature plasma (LTP) is a new green technology, which is widely used in crop seed treatment. Corona plasma is a typical discharge mode of plasma, which can affect the vigor of seeds. The effect of different discharge power on the soybean seed vigor by plasma treatment was experimentally investigated. Plasma discharge characteristic wavelength and spatial distribution were analysed. It shows that the corona discharge spectrum mainly exhibits the strong ultraviolet radiation and 90% of the spectral intensity focused in the center of discharge region. Water absorption and germination index of seeds and the fresh weight of seedlings were used to characterize the specific effects caused by different plasma powers. The results show that plasma treatment has a significant effect on the early stage of germination and can significantly affect the soybean seed vigor and growth. Overdose treatment will cause inhibiting effect. This study provides an experimental basis for the practical agriculture application of corona plasma seed treatment.

Semantic segmentation, as the pixel level classification with dividing an image into multiple blocks based on the similarities and differences of categories (i.e., assigning each pixel in the image to a class label), is an important task in computer vision. Combining RGB and Depth information can improve the performance of semantic segmentation. However, there is still a problem of the way to deeply integrate RGB and Depth. In this paper, we propose a cross-modal feature fusion RGB-D semantic segmentation model based on ConvNeXt, which uses ConvNeXt as the skeleton network and embeds a cross-modal feature fusion module (CMFFM). The CMFFM designs feature channel-wise and spectral-wise fusion, which can realize the deeply feature fusion of RGB and Depth. The in-depth multi-modal feature fusion in multiple stages improves the performance of the model. Experiments are performed on the public dataset of SUN-RGBD, showing the best segmentation by our proposed model ConvNeXt-CMFFM with the highest mIoU score of 53.5% among the nine comparative models. The outstanding performance of ConvNeXt-CMFFM is also achieved on our self-built dataset of RICE-RGBD with the highest mIoU score and pixel accuracy among the three comparative datasets. The ablation experiment on our rice dataset shows that compared with ConvNeXt (without CMFFM), the mIoU score of ConvNext-CMFFM is increased from 71.5% to 74.8% and its pixel accuracy is increased from 86.2% to 88.3%, indicating the effectiveness of the added feature fusion module in improving segmentation performance. This study shows the feasibility of the practical application of the proposed model in agriculture.

Under the background of the Internet of things (IoT), the problems between the actual production and the environment are also prominent. The environmental cost control in the production process of manufacturing enterprises are discussed to reduce the environmental cost and promote the improvement of production efficiency. First, the environmental cost under the background of IoT is analyzed. Also, the environmental cost control methods in the production process of traditional manufacturing enterprises are investigated. Second, based on the principle of traditional genetic algorithm, the fast-nondominated sorting genetic algorithm (NSGA-II) of multiobjective genetic algorithm is introduced to complete the optimization of BP neural network (BPNN) algorithm in deep learning (DL), and the multiobjective GA optimization BPNN model is established. Finally, the multiobjective GA algorithm is used to empirically analyze the environmental cost control capability of a paper-making enterprise. It is compared with enterprises with excellent and poor environmental cost control capabilities in the same industry to find out secondary indexes. The results show that environmental costs have long-term and economic characteristics. The global search ability of BPNN optimized by multiobjective GA is improved, and the local optimal dilemma is avoided. Through empirical analysis, it is found that the comprehensive capability of the environmental cost control of the enterprise is better, scored 79 or more, and the indexes of insufficient development and advantages are obtained. As IoT rapidly develops, it is necessary to further improve the ability of enterprises in environmental cost management, which is very important to promote the development of enterprises and enhance their core competitiveness. It is hoped that this investigation can provide certain reference significance for improving the environmental cost management capability of enterprises, increasing production efficiency, and reducing environmental costs.

As a commonly used plasma diagnostic method, the spectral analysis methodology generates a large amount of data and has a complex quantitative relationship with discharge parameters, which result in low accuracy and time-consuming operation of traditional manual spectral recognition methods. To quickly and efficiently recognize the discharge parameters based on the collected spectral data, a one-dimensional (1D) deep convolutional neural network was constructed, which can learn the data features of different classes of ethylene plasma spectra to obtain the corresponding discharge parameters. The results show that this method has a higher recognition accuracy of higher than 98%. This model provides a new idea for plasma spectral diagnosis and its related application.