Search

The existing non-negative matrix factorization (NMF) algorithms still have some shortcomings. On one hand, the NMF method calculates its low-dimensional representation directly on the high-dimensional original image data set, but in fact the effective information of the original image data set is often hidden in its low-rank structure; on the other hand, the NMF method also has the shortcomings of being sensitive to noise data and unreliable graphs and poor robustness. In order to solve these problems, a non-negative low rank graph embedding (NLGE) algorithm is proposed, which takes into account both the geometric information of the original image data and the effective low-rank structure, and further improves its robustness. In addition, an iteration rule for solving NLGE algorithm is given, and the convergence of the algorithm is further proven. Finally, the experimental results on ORL, CMU PIE, YaleB and USPS databases show the effectiveness of NLGE algorithm.

The structural form of sandwich-structured immersed tunnel (SSIT) can be complex. During the casting of self-compacting concrete, creating void defects between the steel shell and concrete interface is not difficult, which can adversely affect the overall safety and service life of the structure. However, detecting millimeter-scale voids covered by a thick steel plate is a technical challenge for current engineering industries. In this study, we proposed a nondestructive millimeter-scale void detection method for SSITs with thick steel shells by combining impact imaging and neutron methods. First, based on the near-source wavefield theory and count rate of thermal neutrons, the void area and depth calculation methods were derived theoretically, and then the coupling detection method and grading criteria for void severity were proposed. Additionally, the void detection performance was validated for a full-scale SSIT model test by blind detection. Finally, the proposed method was applied to the SSIT of the Shenzhen–Zhongshan bridge. The results showed that the proposed method could quantitatively determine the location and distribution pattern of a void; however, it could not accurately determine the void depth. In contrast, the neutron method could accurately calculate the void depth but had a large minimum detectable unit area. The proposed method could effectively compensate for the limitations of both methods. Statistically, the coincidence rate of the model test was 95%, 89%, and 87.5% for the void location, void area, and void depth, respectively, when the error range was ±2 mm. Using this method, 30 tubes in the Shenzhen–Zhongshan bridge were inspected, and by summarizing the void law, suggestions to improve the casting process were proposed, such as adjusting the casting speed. Meanwhile, the void probability decreased significantly. The proposed method provides an important basis for high-quality construction in SSIT projects. [ABSTRACT FROM AUTHOR]

Air quality monitoring is important to the green development of smart cities. Several technical challenges exist for intelligent, high-precision monitoring, such as computing overhead, area division, and monitoring granularity. In this article, we propose a fine-grained air quality monitoring system based on visual inspection analysis embedded in unmanned aerial vehicle (UAV), referred to as AQMon. This system employs a lightweight neural network to obtain an accurate estimate of atmospheric transmittance in visual information while reducing computation and transmission overhead. Considering that air quality is affected by multiple factors, we design a dynamic fitting approach to model the relationship between scattering coefficients and PM2.5 concentration in real time. The proposed system is evaluated using public datasets and the results show that AQMon outperforms four existing methods with a processing time of 13.8 ms. [ABSTRACT FROM AUTHOR]

A middle-low-speed maglev train is supported by an electromagnetic force between the suspension electromagnet (EM) and the steel rail and is driven by a linear induction motor. The capability of the suspension system has a direct bearing on safety and the technical and economic performance of the train. This paper focuses on the dependence of the electromagnetic force on the structural configuration of the EM with the purpose of improving performance of a conventional EM. Finally, a novel configuration is proposed of a hybrid suspension magnet, which combines permanent magnets and coils, in order to increase the suspension force while reducing the suspension power loss.

The genus Quambalaria includes several important pathogens of species of Eucalyptus and Corymbia, mainly causing leaf and shoot blight. Recently, extensive shoot and leaf dieback and stem cankers suspected to be Quambalaria diseases have been found on young Eucalyptus urophylla × E. grandis trees in Guangdong and Hainan Provinces. The occurrence of Quambalaria species and their association with eucalypt hosts within China needs to be investigated for tree diseases management. The isolates from the diseased samples were identified based on their morphological structures and phylogenetic analyses with DNA sequence data for the internal transcribed spacer region and large ribosome subunit RNA of the nuclear rDNA. This work revealed that three species of Quambalaria were present: Quambalaria pitereka from Corymbia citriodora, Q. eucalypti from E. urophylla × E. grandis, both isolated from young eucalypt leaves and shoots in Guangdong Province, and Quambalaria simpsonii, which was isolated from stem cankers of E. urophylla × E. grandis at four different sites across Guangdong and Hainan Provinces. These results confirmed that Quambalaria agents were associated with the diseases occurring on eucalypt hosts in South China. This is the first report of Q. eucalypti in Asia and the first report of Q. simpsonii in China on Eucalyptus trees.

To explore the coupling of dry-wet seasonal variations of soil respiration with their environmental factors in the alpine meadow under the background of increasing nitrogen (N) deposition, we conducted an experiment in the typical degraded Poa pratensis meadow in the Napahai, Yunnan. There were four treatments, i.e., control (0 g • m-2 • a-2), low (5 g • m-2 • a-2), medium (10 g • m-2 • a-2), and high (15 g • m-2 • a-2) levels. We examined the effects of aboveground biomass, plant diversity, and soil physicochemical properties on soil respiration. The results showed that N deposition significantly promoted soil respiration. Compared with that in the control, soil respiration rates increased by 21.9%-53.9% and 27.3%-51.2% in dry and wet seasons, respectively. The maximum value of soil respiration rate was recorded in the medium N treatment. N deposition dramatically elevated aboveground biomass (52.2%-66.4%). Plant diversity declined with increasing N addition levels, with the maximum value (13.5%-24.2%) being recorded in high treatment in wet season. The values of ammonium nitrogen, organic matter, microbial biomass carbon and nitrogen, temperature and moisture in the three N treatments were elevated by 14.3%-333.5% compared with the control, while those of soil pH were decreased by 9.0%-34.6%. Results of the structural equation modelling showed that plant biomass, Shannon diversity, microbial biomass, soil temperature, and moisture showed a positive effect on soil respiration, while bulk density had a negative effect. Soil nitrogen pool and pH were main factors driving soil CO2 emissions, accounting for 55.7% and 45.1% of the variations, respectively. Therefore, short-term atmospheric N deposition stimulated soil respiration primarily via altering soil pH and nitrogen pool components in the degraded alpine meadow. [ABSTRACT FROM AUTHOR]