Search

Forest fires have continually endangered personal safety and social property. To reduce the occurrences of forest fires, it is essential to detect forest fire smoke accurately and quickly. Traditional forest fire smoke detection based on convolutional neural networks (CNNs) needs many hand-designed components and shows poor ability to detect small and inconspicuous smoke in complex forest scenes. Therefore, we propose an improved early forest fire smoke detection model based on deformable transformer for end-to-end object detection (deformable DETR). We use deformable DETR as a baseline containing the best sparse spatial sampling for smoke with deformable convolution and relation modeling capability of the transformer. We integrate a Multi-scale Context Contrasted Local Feature module (MCCL) and a Dense Pyramid Pooling module (DPPM) into the feature extraction module for perceiving features of small or inconspicuous smoke. To improve detection accuracy and reduce false and missed detections, we propose an iterative bounding box combination method to generate precise bounding boxes which can cover the entire smoke object. In addition, we evaluate the proposed approach using a quantitative and qualitative self-made forest fire smoke dataset, which includes forest fire smoke images of different scales. Extensive experiments show that our improved model’s forest fire smoke detection accuracy is significantly higher than that of the mainstream models. Compared with deformable DETR, our model shows better performance with improvement of mAP (mean average precision) by 4.2%, APS (AP for small objects) by 5.1%, and other metrics by 2% to 3%. Our model is adequate for early forest fire smoke detection with high detection accuracy of different-scale smoke objects.

Deep learning-based forest fire vision monitoring methods have developed rapidly and are becoming mainstream. The existing methods, however, are based on enormous amounts of data, and have issues with weak feature extraction, poor small target recognition and many missed and false detections in complex forest scenes. In order to solve these problems, we proposed a multi-task learning-based forest fire detection model (MTL-FFDet), which contains three tasks (the detection task, the segmentation task and the classification task) and shares the feature extraction module. In addition, to improve detection accuracy and decrease missed and false detections, we proposed the joint multi-task non-maximum suppression (NMS) processing algorithm that fully utilizes the advantages of each task. Furthermore, considering the objective fact that divided flame targets in an image are still flame targets, our proposed data augmentation strategy of a diagonal swap of random origin is a good remedy for the poor detection effect caused by small fire targets. Experiments showed that our model outperforms YOLOv5-s in terms of mAP (mean average precision) by 3.2%, APS (average precision for small objects) by 4.8%, ARS (average recall for small objects) by 4.0%, and other metrics by 1% to 2%. Finally, the visualization analysis showed that our multi-task model can focus on the target region better than the single-task model during feature extraction, with superior extraction ability.

A novel co-catalyst system under visible-light irradiation was constructed using high-purity metal and alloy mesh and a Mn0.5Cd0.5S photocatalyst with a narrow band gap (1.91 eV) prepared by hydrothermal synthesis. The hydrogen production rate of Mn0.5Cd0.5S changed from 2.21 to 6.63 mmol·(g·h)−1 with the amount of thioacetamide, which was used as the sulphur source. The introduction of Ag, Mo, Ni, Cu, and Cu–Ni alloy meshes efficiently improved the H2 production rate of the co-catalyst system, especially for the Ni mesh. The improvement can reach an approximately six times greater production, with the highest H2 production rate being 37.65 mmol·(g·h)−1. The results showed that some bulk non-noble metal meshes can act as good or better than some noble metal nanoparticles deposited on the main photocatalyst for H2 evolution due to the promotion of photoinduced electron transfer, increase in redox reaction sites, and prevention of the recombination of carriers.

In the big data environment, data are characterized by a large volume, various types, and rapid changes. The consortium blockchain applied in this environment faces the problem of excessive storage of the ledger, and the ledger handling different types of business needs to be isolated to ensure the ledger’s security. To this end, this paper proposes a scalable consortium blockchain architecture based on world state collaborative storage (CBCS). First, a business world state database update method is designed based on sparse Merkle multiproofs, where the collaborative storage of world state is realized under the premise of mutual isolation of the ledger between business domains. Then, a world state consistency verification method based on the rank B+ tree is designed to verify the consistency of the business world state in business domains by the checking sidechain, and a main-side chain cross-anchoring structure is designed to realize secure anchoring of the mainchain and the checking sidechain. Meanwhile, a blockchain transaction trusted tracing method based on two-level certification is developed to enable business nodes to obtain complete blockchain transactions. Finally, the feasibility and efficiency of the proposed mechanism to solve the storage scalability problem in the consortium blockchain are verified through experiments.

Biochar soil amendment has been increasingly recommended for enhancing soil fertility and crop productivity while reducing greenhouse gas emissions in agricultural soils. However, a clear understanding of the cost benefits and longevity of the positive effects over long term would be a prerequisite for large scale biochar production and application in agriculture. In this study, the long-term effects of a single biochar amendment on soil fertility, crop yield and greenhouse gas emissions were assessed in a six-year field experiment of a fertile rice paddy from Southwest China. The field trial was established in 2010 and was managed under a rice-winter wheat rotation system throughout six years until 2016. The experiment employed a nested design with biochar soil amendment at application rates of 0, 20 and 40 t ha−1 without N fertilization, and 150 kg ha−1 with N fertilization, respectively. Soil properties and crop yields were measured and the emissions of CH4 and N2O were monitored during the rice cultivation period in the 1st, 2nd and 5th year since 2010. Soil pH, organic carbon, total N and available potassium content were all increased under biochar amendment and persisted throughout the six years. However, grain yields across the years were generally not affected by biochar amendment except a yield increase in 2015. This grain yield increase (18.3%) was reported in the wheat season of 2015, when biochar had been applied at 40 t ha−1 in plots along with N fertilizer. Biochar amendment reduced N2O emissions from the rice paddies only for the first two seasons following the single amendment. In conclusion, a single but high rate of biochar amendment provided limited and temporary benefits for improving grain yields and reduction of greenhouse gases in the fertile paddy soil.

In this study, hydrocracking of Gudao vacuum residue was carried out in an internally stirred autoclave in the presence of several kinds of dispersed water and oil soluble unsupported Mo catalysts. The benzene insolubles formed were examined by X-ray diffraction (XRD), scanning electron microscopy (SEM) and transmission electron microscopy (TEM). The chemical compositions of the cracked gases and residues were analysed by gas chromatography and liquid chromatography (SARA method), respectively. From the XRD analyses, it was identified that the possible catalytically active phase generated from the dispersed Mo catalyst precursor was microcrystalline MoS2. The results indicate that the high dispersion method of the water soluble catalyst precursor postulated in this study, can promote catalytic activity of the catalyst. The experimental results also suggest that the dispersed-phase Mo catalyst can considerably inhibit polycondensation of resins and asphaltenes, and delay coke formation by way of promoting hydrogenation and hydrodesulfurization.