Fine-grained image classification on bats using VGG16-CBAM: a practical example with 7 horseshoe bats taxa (CHIROPTERA: Rhinolophidae: Rhinolophus) from Southern China.

BACKGROUND: Rapid identification and classification of bats are critical for practical applications. However, species identification of bats is a typically detrimental and time-consuming manual task that depends on taxonomists and well-trained experts. Deep Convolutional Neural Networks (DCNNs) provide a practical approach for the extraction of the visual features and classification of objects, with potential application for bat classification. RESULTS: In this study, we investigated the capability of deep learning models to classify 7 horseshoe bat taxa (CHIROPTERA: Rhinolophus) from Southern China. We constructed an image dataset of 879 front, oblique, and lateral targeted facial images of live individuals collected during surveys between 2012 and 2021. All images were taken using a standard photograph protocol and setting aimed at enhancing the effectiveness of the DCNNs classification. The results demonstrated that our customized VGG16-CBAM model achieved up to 92.15% classification accuracy with better performance than other mainstream models. Furthermore, the Grad-CAM visualization reveals that the model pays more attention to the taxonomic key regions in the decision-making process, and these regions are often preferred by bat taxonomists for the classification of horseshoe bats, corroborating the validity of our methods. CONCLUSION: Our finding will inspire further research on image-based automatic classification of chiropteran species for early detection and potential application in taxonomy.

Detection of Multi-Layered Bond Delamination Defects Based on Full Waveform Inversion.

This study aimed to address the challenges encountered in traditional bulk wave delamination detection methods characterized by low detection efficiency. Additionally, the limitations of guided wave delamination detection methods were addressed, particularly those utilizing reflected waves, which are susceptible to edge reflections, thus complicating effective defect extraction. Leveraging the full waveform inversion algorithm, an innovative approach was established for detecting delamination defects in multi-layered structures using ultrasonic guided wave arrays. First, finite element modeling was employed to simulate guided wave data acquisition by a circular array within an aluminum-epoxy bilayer structure with embedded delamination defects. Subsequently, the full waveform inversion algorithm was applied to reconstruct both regular and irregular delamination defects. Analysis results indicated the efficacy of the proposed approach in accurately identifying delamination defects of varying shapes. Furthermore, an experimental platform for guided wave delamination defect detection was established, and experiments were conducted on a steel-cement bilayer structure containing an irregular delamination defect. The experimental results validated the exceptional imaging precision of our proposed technique for identifying delamination defects in multi-layered boards. In summary, the proposed method can accurately determine both the positions and sizes of defects with higher detection efficiency than traditional pulse-echo delamination detection methods.

Multi-perspective evaluation on spent lithium iron phosphate recycling process: For next-generation technology option.

The recycling of spent lithium iron phosphate batteries has recently become a focus topic. Consequently, evaluating different spent lithium iron phosphate recycling processes becomes necessary for industrial development. Here, based on multiple perspectives of environment, economy and technology, four typical spent lithium iron phosphate recovery processes (Hydro-A: hydrometallurgical total leaching recovery process; Hydro-B(H2O2/O2): hydrometallurgical selective lithium extraction process; Pyro: Pyrometallurgical recovery process; Direct: Direct regeneration process) were compared comprehensively. The comprehensive evaluation study uses environment, economy and technology as evaluation indicators, and uses the entropy weight method and analytic hierarchy process to couple the comprehensive indicator weights. Results show that the comprehensive evaluation values of Hydro-A, Hydro-B (H2O2), Hydro-B (O2), Pyro and Direct are 0.347, 0.421, 0.442, 0.099 and 0.857, respectively. Therefore, the technological maturity of Direct should be further improved to enable early industrialization. On this basis, this study conducted a quantitative evaluation of the spent lithium iron phosphate recycling process by comprehensively considering environmental, economic and technical factors, providing further guidance for the formulation of recycling processes.

High-Precision Corrosion Detection via SH1 Guided Wave Based on Full Waveform Inversion.

Corrosion detection for industrial settings is crucial for safe and efficient operations. Due to its high imaging resolution, the guided-wave full-waveform inversion tomography technique has significant potential for corrosion detection of plate metals. Limited by the long wavelengths of A0 and S0 mode waves, this method exhibits inadequate detection resolution for the earlier shallow and small corrosion defects. Based on the relatively short wavelength characteristics of the SH1 mode wave, we propose a high-precision corrosion detection method via SH1 guided wave using the full waveform inversion algorithms. By conducting finite element simulations of ultrasonic-guided waves on aluminum plates with varying corrosion defects, a comparison was made to assess the detection precision across A0, S0, and SH1 modes. The comparison results showed that, whether for regular or irregular defects, the SH1 mode wave always exhibited higher imaging accuracy than the A0 and S0 mode waves for shallow and small-sized defects. The corresponding experiments were conducted on an aluminum plate with simple or complex defects. The results of the experiments reconfirmed that the full waveform inversion method using SH1 guided wave can effectively reconstruct the shape and size of small and shallow corrosion defects within aluminum plates.

Mechanism of highly efficient oil removal from spent hydrodesulfurization catalysts by ultrasound-assisted surfactant cleaning methods.

The removal of crude oil from spent hydrodesulfurization catalysts constitutes the preliminary stage in the recovery process of valuable metals. However, the traditional roasting method for the removal exhibits massive limitations. In view of this, the present study used an ultrasound-assisted surfactant cleaning method to remove crude oil from spent hydrodesulfurization catalysts, which demonstrated effectiveness. Furthermore, the study investigated the mechanism governing the process with calculation and experiments, so as to provide a comprehensive understanding of the cleaning method's efficacy. The surfactant selection was predicated on the performance in the IFT test, with SDBS and TX-100 finally being chosen. Subsequent calculations and analysis were then conducted to elucidate their frontier molecular orbitals, electrostatic potential, and polarity. It has been found that both SDBS and TX-100 possess the smallest LUMO-HOMO energy gap (ΔE), registering at 4.91 eV and 4.80 eV, respectively, and presenting the highest interfacial reactivity. The hydrophilic structure in the surfactant regulates the wettability of the oil-water interface, and the long-chain alkanes have excellent non-polar properties that promote the dissolution of crude oil. The ultrasonic-assisted process further improves the interface properties and enhances the oil removal effect. Surprisingly, the crude oil residue was reduced to 0.25% under optimal conditions. The final phase entailed the techno-economic evaluation of the entire process, revealing that, in comparison to the roasting method, this process saves $0.38 per kilogram of spent HDS catalyst, with the advantages of operational simplicity and emission-free. Generally, this study shed new light on the realization of efficient oil removal, with the salience of green, sustainable, and economical.

[Influence of Land Use Structure and Spatial Pattern on Water Quality of Small and Medium-sized Rivers in Poyang Lake Basin].

To reveal the influence mechanism of land use structure and spatial pattern on water quality of small and medium-sized rivers, water samples were collected from 25 sampling points in three small and medium-sized rivers of the Poyang Lake Basin in January 2022 and July 2022. Bioenv analysis, the Mantel test, and variance partitioning analysis were used to quantify the effects of land use structure and spatial patterns on water quality at different spatial scales; generalized additive models were used to fit the relationship between water quality and different land use structures and spatial patterns; and a generalized linear model was used to construct segmented regression models and calculate the thresholds based on the stepwise recursive method. The results showed that:① the average interpretation rate of land use structure and spatial pattern on river water quality was 59.72% during the wet period and 48.95% during the dry period. The sub-basin and riparian 100 m scales were the key scales of land use structure and spatial pattern affecting water quality in small and medium-sized rivers, with an average explanation rate of 54.70% and 64.88%, respectively. The joint explanation of land use structure and spatial pattern was an important factor driving the change in river water quality, accounting for 66.90% of the total explanation. ② The impact of land use structure on the water quality of small and medium-sized rivers had a significant threshold effect. When the proportion of construction land was less than 2%, farmland was less than 8%, or forest land was more than 82% at the sub-basin scale and the proportion of construction land was less than 12%, farmland was less than 41%, or forest land was more than 49% at the riparian buffer scale, all could significantly improve water quality. ③ The effect of spatial pattern on water quality in small and medium-sized rivers also had a threshold effect but was weaker than that of land use structure. A patch shape value more than 28.77 or patch diversity more than 0.69 at the sub-basin scale and a patch shape value more than 2.99 or patch diversity more than 1.02 at the riparian buffer scale could improve water quality. The above results showed that strengthening the management of land use at the sub-basin and riparian 100 m scales and setting a reasonable threshold of land use structure and spatial pattern can effectively prevent water quality from deteriorating.