Infrared Physics & Technology最新文献

The continuous mining activities at the working face cause periodic disturbances to the surrounding coal rock mass. Investigating the response characteristics of surface infrared radiation temperature, energy evolution features, and fracture development patterns in coal rock under cyclic loading and unloading is crucial for enhancing our understanding of the mechanisms of damage, degradation, and instability in rocks associated with mining operations. In this research work, we conducted infrared radiation observation experiments on both dry and wet coal samples subjected to uniaxial cyclic loading and unloading. Furthermore, we carried out a thorough analysis of the resulting temperature variations, energy changes, and fracture evolution patterns. It was found that the average infrared radiation temperature (AIRT) generally increases during the loading phase and decreases during the unloading phase, with the variance of successive minus infrared image temperature (VSMIT) exhibiting a sharp change just prior to failure. Energy analysis indicates that dissipated energy during the pore compaction and elastic deformation stages is minimal, while in the plastic deformation stage, the proportion of dissipated energy to total energy increases, displaying a “concave” upward trend. Additionally, fitting results show that the AIRT follows a single exponential decay relationship with dissipated energy, with wet coal exhibiting a greater decay coefficient, highlighting that moisture accelerates the rate of temperature decline. The Particle Flow Code (PFC) simulation results further demonstrate that the number of cracks in dry coal samples significantly exceeds that in wet coal samples, showing a single exponential relationship between the number of fractures and dissipated energy, which indicates that the development of fractures in dry coal rock occurs at a faster rate with increasing dissipated energy compared to wet coal rock.

A 100 Hz compact 2 μm optical parametric oscillator (OPO) based on a type II non-critically phase-matched KTiOPO4 crystal (KTP) was reported. The monolithic KTP crystal was pumped with a passively Q-switched 1064 nm Nd:YAG laser. At a repetition rate of 100 Hz and a single pulse energy of 5.8 mJ of pump light, a 2128-nm laser output was achieved with a maximum of 1.17 mJ of parametric light at the degenerate point, with a parametric light pulse width of approximately 10.5 ns, corresponding to a pump to parametric light conversion efficiency of 20.1 %.

Flavonoids are compounds resulting from secondary plant metabolism that provide benefits to human health by food. This study aimed to accuracy of predicting flavonoids in maize plants subjected to different nitrogen rates using hyperspectral reflectance and machine learning (ML) algorithms. The experiment was carried out in randomized blocks in a 4 × 5 factorial design (four N inputs: 0; 30; 60 and 120 % of the recommended N input; and five readings of the reflectance spectra in maize leaves from different vegetative stages: V6, V8, V10, V12 and V14, in four replications, totaling 80 treatments. N rates were applied in the V4 and V8 phenological stages, using urea as the N source. For hyperspectral analysis, four leaves from each treatment were collected and analyzed using a spectroradiometer (FieldSpec 4 HRes, Analytical Spectral Devices), capturing the spectrum in the 350 to 2500 nm range. Subsequently, the leaf samples used in the reflectance readings were dried, ground and subjected to isoflavone quantification, analyzed by ultra-performance liquid chromatography in three repetitions, quantifying daidzein 1 (D1), daidzein 2 (D2), genistein 1 (G1), genistein 2 (G2), and total isoflavones. Data obtained was subjected to machine learning analysis, testing two data set input configurations: wavelengths (WL) and calculated spectral bands (B), and D1, D2, G1, G2 and total isoflavones as output variables. The ML algorithms tested were artificial neural networks (ANN), REPTree (DT), M5P decision tree (M5P), ZeroR (R), Random Forest (RF) and support vector machine (SVM), evaluated according to their performance by the correlation coefficient (r) and mean absolute error (MAE). The results show that the SVM algorithm had the highest accuracy in predicting the variables D1, D2, G1, G2 and total isoflavones, outperforming the other algorithms when WL was used as input in dataset.

Problem

In many countries, agriculture is the main source of people’s livelihood and satisfies their nutritional needs. Early detection of plant diseases through agricultural remote monitoring is important to prevent the disease’s spread. The traditional methods require sampling and can damage the plant, but hyperspectral imaging is non-destructive.

Aim

The major aim of this research is to devise a Water Wheel Plant Dingo Optimizer_Deep Convolutional Neural Network (WWPDO_Deep CNN) for disease detection using a hyperspectral leaf image.

Methods

Initially, the input leaf image is given into the leaf segmentation phase, which is done using the proposed Water Wheel Plant Dingo Optimizer (WWPDO), which is the amalgamation of the Water Wheel Plant Algorithm (WWPA) and Dingo Optimizer (DOX). The selected bands’ outputs are subjected to leaf segmentation and which is carried out by employing Bayesian Fuzzy Clustering (BFC). Thereafter, leaf segmented outputs are fussed using the majority voting method. Fused output and individual leaf segmentation output are given into the feature extraction process to extract features, such as local binary patterns and Weber local descriptors. Finally, leaf disease detection is performed using a deep Convolutional Neural Network (Deep CNN) for normal and abnormal cases. The hyperparameters of the Deep CNN are fine-tuned based on the proposed WWPADO.

Results

The proposed WWPDO_Deep CNN achieved an excellent performance with an accuracy of 91.35 %, a True Positive Rate (TPR) of 93.13 % and a True Negative Rate (TNR) of 90.76 %.

Conclusion

The WWPDO_Deep CNN is applicable for early diagnosis under the new classification system and provides a new direction for early diagnosis based on hyperspectral images. Also, the devised model provides an accurate diagnosis of plant diseases. Early and accurate detection allows targeted treatment, reduces the need for widespread pesticide application and promotes more sustainable farming practices.

Different wheat flour types are used to produce various baked products. Due to the whiteness of the four types, hyperspectral imaging can be used due to receiving infrared wavelength. The technique was applied to distinguish confectionery flour and the flours of Samoun, Sangak, and Tafton breads using a line scanning system in the range of 400–950 nm. Effective wavelengths were selected and different image features were extracted from the corresponding image channels. The selected wavelengths were 601.33, 620.34, 696.41, 730.31, 821.26, and 841.11 nm. The extracted features were used in classification step using linear discriminant analysis, support vector machine, and artificial neural network methods in MATLAB software. The classification accuracy of artificial neural network was higher than the other methods. The efficient features gave higher classification accuracy (98.1 %) than all extracted features (96.9 %). The results showed the high ability of hyperspectral imaging combined with artificial neural network to distinguish different wheat flour types.

Grape shelf-life estimation is a substantial challenge for the grape industry. The objective of this study is to investigate the potential of grape shelf-life estimation using HSI technique and a deep learning algorithm. The visible and near-infrared (400.68–1001.61 nm) hyperspectral reflectance images data of grape samples was acquired and preprocessed with different spectral preprocessing methods. Additionally, a stacked denoising autoencoder (SDAE)-based deep learning algorithm was developed to extract deep features from pixel-level hyperspectral data of grapes, and then these features were used as inputs to establish support vector machine (SVM) models for estimating grape shelf-life. Furthermore, SVM, one-dimensional convolutional neural network (1D CNN) and long short-term memory (LSTM) models were used as traditional machine learning and end to end models for comparison. The results demonstrated that the SDAE-SVM model achieved reasonable recognition accuracy of 100 % and 98.125 % for the shelf-life of grapes in the training and test sets, respectively. The overall results suggested that SDAE-based deep learning method can be used as a powerful tool to deal with large-scale hyperspectral data as well as this research confirms the feasibility of non-destructive estimation for grapes shelf-life by the combination of HSI technique and deep learning method, which would provide a valuable guidance for shelf-life estimation of other postharvest fruit.

Due to the strong time-varying characteristics and complex geometry of aerospace components, rapid changes in the distribution of radiative heat flux on the test surface are often required during non-uniform aerodynamic heating tests. However, it takes time to adjust the geometric parameters of the lamp array, which cannot meet the requirement for rapidly changing radiative heat flux distribution. To address this issue, a new method for calculating radiant heat flux and a fast linear analysis method of quartz lamp power are proposed which can calculate radiant heat flux distribution of complex surface and meet the need of timeliness and rapidity in radiant heat flux distribution, making it more suitable for engineering applications. Through numerical verification under single lamp and quartz lamp array, the maximum difference between the theoretical analysis method and the Monte Carlo method is less than 2.25% under single lamp, and less than 5% under quartz lamp array. Finally, turbine blade model and plane model are taken as research objects to verify the feasibility and reliability of the fast linear analysis method of quartz lamp power. The results show that the relative average error of the calculated quartz lamp power is 5.86% and 14.83%, respectively, compared with the actual power. This provides a reference and basis for the rapid simulation design of thermal radiation environment during the experiment.

Metalenses have a high design degree of freedom in controlling the light field and excellent performance in chromatic aberration elimination. However, in designing ultra-broadband achromatic metalenses, integrating multiple types of unit structures is necessary to compensate for phase differences caused by different incident wavelengths. Here, we propose an ultra-broadband achromatic metalens composed only of a single type of square nano-pillar. which controls the entire operating band by utilizing three wide-band fusions, and have characteristics depending on the phase coverage and relative phase. In the operational range of 2–5 μm, the achromatic metalens demonstrates a maximum focal shift of 2.1 μm. The average focal shift is 0.98 %, the average NA value is 0.35, with an average relative phase of 0.77π. The average transmittance and focus efficiency are 94.17 % and 58.7 %, respectively. This broad-spectrum fusion design strategy simplifies manufacturing complexity while maintaining high focusing efficiency and transmittance levels throughout the entire operational bandwidth. This design approach can improve image resolution and quality by minimizing chromatic aberration.

Infrared thermal radiation emission in the 8.5 – 28 THz frequency region is obtained using surface metal–semiconductor grating structures on undoped (u-) GaAs, u-GaP, u-ZnO, u-GaN, and n-type SiC in a temperature range of 430 – 630 K. These emissions resonate with longitudinal optical (LO) phonon or LO-like lattice vibration energies determined by the zero points of the real parts of the dielectric functions in the surface structures. The emissions of materials with Reststrahlen bandwidths of a few tens of cm⁻¹ show the emissions resonating with their LO phonon modes, while materials with bandwidth of more than 170 cm⁻¹ show peak energies significantly lower than the LO phonon: LO-like phonon resonance. The emission intensity is found to be dominated by the balance of radiative and nonradiative LO or LO-like phonon annihilation rates. The radiative rate is dominated by the LO-phonon–radiation interaction Hamiltonian and the Bose-Einstein factor. High emission intensity is obtained for the structure on u-ZnO with intense LO-like phonon–radiation interaction. The dependence of the emission intensity on temperature and emission window width for various materials shows the effect of material-dependent metal/semiconductor interface conditions on the emission efficiency.