Visible and NIR microscopic hyperspectrum reconstruction from RGB images with deep convolutional neural networks.

We investigate the microscopic hyperspectral reconstruction from RGB images with a deep convolutional neural network (DCNN) in this paper. Based on the microscopic hyperspectral imaging system, a homemade dataset consisted of microscopic hyperspectral and RGB image pairs is constructed. For considering the importance of spectral correlation between neighbor spectral bands in microscopic hyperspectrum reconstruction, the 2D convolution is replaced by 3D convolution in the DCNN framework, and a metric (weight factor) used to evaluate the performance reconstructed hyperspectrum is also introduced into the loss function used in training. The effects of the dimension of convolution kernel and the weight factor in the loss function on the performance of the reconstruction model are studied. The overall results indicate that our model can show better performance than the traditional models applied to reconstruct the hyperspectral images based on DCNN for the public and the homemade microscopic datasets. In addition, we furthermore explore the microscopic hyperspectrum reconstruction from RGB images in infrared region, and the results show that the model proposed in this paper has great potential to expand the reconstructed hyperspectrum wavelength range from the visible to near infrared bands.

Ultranarrow-bandwidth planar hot electron photodetector based on coupled dual Tamm plasmons.

Hot electron photodetectors based on a planar structure of metal-insulator /semiconductor-metal (MIM/MSM) have attracted much attention due to the easy and cheap fabrication process and the possibility of detecting light with energy lower than the semiconductor band gap. For this type of device, however, hot electron photocurrent is restricted by the trade-off between the light absorption and the internal quantum efficiency (IQE) since high absorption usually occurs within thick metals and the IQE in this case is usually low. The trade-off is circumvented in this paper by proposing a new type of hot electron photodetector based on planar MIM structure and coupled dual Tamm plasmons (TPs), which has a structure of one-dimensional photonic crystals (1DPCs)/Au/TiO2/Au/1DPCs. The coupled modes of the dual TPs at the two 1DPCs/Au interfaces can lead to a high absorption of 98% in a 5 nm-thick Au layer. As a result, the responsivity of the conventional device with two Schottky junctions in series configuration reaches a high value of 9.78 mA/W at the wavelength of 800 nm. To further improve the device performance, devices with four Schottky junctions in parallel configuration are proposed to circumvent the hot electrons loss at the interface of the Au layer and the first TiO2 layer of the 1DPCs. Correspondingly, the hot electrons photocurrent doubles and reaches a higher value of 21.87 mA/W. Moreover, the bandwidth of the responsivity is less than 0.4 nm, the narrowest one when compared with that for the hot electron photodetectors reported so far in the published papers.

Tunable optical forces enhanced by plasmonic modes hybridization in optical trapping of gold nanorods with plasmonic nanocavity.

The optomechanical interaction between a plasmonic nanocavity and a gold nanorod through optical forces is demonstrated. It is revealed that strong localized plasmon resonance mode hybridization induced by a gold nanorod results in the resonance mode of the nanocavity splitting into two different plasmon resonance modes (bonding plasmon resonance mode and antibonding plasmon resonance mode). When the whole system (gold nanorod and gold nanocavity) is excited at the antibonding plasmon mode, the gold nanorod can receive an optical pushing force and be pushed away from the gold nanocavity. On the other hand, an optical pulling force acts on the gold nanorod and the gold nanorod can be trapped by the gold nanocavity when the plasmonic tweezers work at the bonding mode. The optical pulling force acting on the gold nanorod can be enhanced by two orders of magnitude larger than that of the same sized dielectric nanorod, which benefits from the strong resonant nearfield interaction between the gold nanorod and the gold nanocavity. More importantly, the shape and the position of the optical potential can be tuned by tailoring the wavelength of the laser used in the optical trapping, which can be used to manipulate the gold nanorod within a nanoscale region. Our findings have important implications for optical trapping, manipulation, sorting, and sieving of plasmonic nanoparticles with plasmonic tweezers.

Improving the charge carrier transport of organic solar cells by incorporating a deep energy level molecule.

Tetrafluoro-tetracyanoquinodimethane (F4-TCNQ), a strong molecular acceptor, has been proved to be an excellent candidate to achieve the p-type doping effect. When F4-TCNQ is incorporated into a poly(3-hexylthiophene) (P3HT): indene-C60 bisadduct (ICBA) active layer, superior behavior upon inducing polymer donor excited electron transport is demonstrated due to the addition of a deep-lying lowest unoccupied molecular orbital (LUMO) from F4-TCNQ, leading to the realization of organic solar cells (OSCs) with an improved power conversion efficiency (PCE) of 5.83%, accounting for 29.6% enhancement. In the system of active layer, the low LUMO of F4-TCNQ can easily accept electrons, remarkably reducing electron/hole recombination, which contributes to the enhancement of the photoconductivity and charge carrier mobility, resulting in higher short-circuit current density (Jsc), and achieving a more balanced charge carrier transport, as well as an ideal fill factor (FF).

Semitransparent polymer solar cells with simultaneously improved efficiency and color rendering index.

Herein, we demonstrate a kind of high performance semi-transparent polymer solar cell (STPSC) with a significantly improved color rendering index (CRI) and power conversion efficiency (PCE) by introducing one-dimensional photonic crystals (1DPCs), which are intentionally designed to strongly reflect the pristine weak absorbed light to flatten the concavo-convex transmittance spectrum of STPSCs. The transmitted light from the STPSC device with 4 pairs of 1DPCs under AM 1.5G illumination shows extraordinary color rendering capacities, which contribute an increased CRI from 79 to 91, combined with an enhanced PCE from 4.14% to 5.01% compared to devices without 1DPCs. The simultaneously improved optical and electrical performance suggests that STPSCs can provide a unique feature, which is suitable for building integrated photovoltaic applications.

Explanatory Object Part Aggregation for Zero-Shot Learning.

Zero-shot learning (ZSL) aims to recognize objects from unseen classes only based on labeled images from seen classes. Most existing ZSL methods focus on optimizing feature spaces or generating visual features of unseen classes, both in conventional ZSL and generalized zero-shot learning (GZSL). However, since the learned feature spaces are suboptimal, there exists many virtual connections where visual features and semantic attributes are not corresponding to each other. To reduce virtual connections, in this paper, we propose to discover comprehensive and fine-grained object parts by building explanatory graphs based on convolutional feature maps, then aggregate object parts to train a part-net to obtain prediction results. Since the aggregated object parts contain comprehensive visual features for activating semantic attributes, the virtual connections can be reduced by a large extent. Since part-net aims to extract local fine-grained visual features, some attributes related to global structures are ignored. To take advantage of both local and global visual features, we design a feature distiller to distill local features into a master-net which aims to extract global features. The experimental results on AWA2, CUB, FLO, and SUN dataset demonstrate that our proposed method obviously outperforms the state-of-the-arts in both conventional ZSL and GZSL tasks.

Detecting bioactive compound contents in Dancong tea using VNIR-SWIR hyperspectral imaging and KRR model with a refined feature wavelength method.

Hyperspectral imaging (HSI) provides opportunity for non-destructively detecting bioactive compounds contents of tea leaves and high detection accuracy require extracting effective features from the complex hyperspectral data. In this paper, we proposed a feature wavelength refinement method called interval band selecting-competitive adaptive reweighted sampling-fusing (IBS-CARS-Fusing) to extract feature wavelengths from visible-near-infrared (VNIR) and short-wave-near-infrared (SWIR) hyperspectral images. Combined with the proposed IBS-CARS-Fusing method, a kernel ridge regression (KRR) model was established to predict the contents of bioactive compounds including chlorophyll a, chlorophyll b, carotenoids, tea polyphenols, and amino acids in Dancong tea. It was revealed that the IBS-CARS-Fusing method can improve Rp2 of KRR model for these bioactive compounds by 4.77%, 4.60%, 6.74%, 15.52%, and 13.10%, respectively, and Rp2 of the model reached high values of 0.9500, 0.9481, 0.8946, 0.8882, and 0.8622. Additionally, a leaf compound mass per area thermal map was used to visualize the spatial distribution of the compounds.

Preparation of nano-polycrystalline WO3 thin films and their solid-state electrochromic display devices.

In this paper, nano-polycrystalline WO3 thin films with the thickness in the range of 100-200 nm have been uniformly prepared on the designed regions of ITO (indium tin oxide) glass substrates by thermal evaporation deposition. Their crystal structures, surface morphologies and uniformities are investigated by X-ray diffraction (XRD), scanning electron microscopy (SEM), and atomic force microscopy (AFM), respectively. The solid-state electrochromic display (ECD) devices based on these nano-polycrystalline WO3 thin films have been also fabricated and have demonstrated to have better performance than normal thin films, including shorter response time, higher contrast, and furthermore, higher stability to keep the colored state without power consumption. These results demonstrate nano-polycrystalline WO3 thin films can be applied to improve the performance of ECD devices, especially suitable to static display.

Detection of Rice Pests Based on Self-Attention Mechanism and Multi-Scale Feature Fusion.

In recent years, the occurrence of rice pests has been increasing, which has greatly affected the yield of rice in many parts of the world. The prevention and cure of rice pests is urgent. Aiming at the problems of the small appearance difference and large size change of various pests, a deep neural network named YOLO-GBS is proposed in this paper for detecting and classifying pests from digital images. Based on YOLOv5s, one more detection head is added to expand the detection scale range, the global context (GC) attention mechanism is integrated to find targets in complex backgrounds, PANet is replaced by BiFPN network to improve the feature fusion effect, and Swin Transformer is introduced to take full advantage of the self-attention mechanism of global contextual information. Results from experiments on our insect dataset containing Crambidae, Noctuidae, Ephydridae, and Delphacidae showed that the average mAP of the proposed model is up to 79.8%, which is 5.4% higher than that of YOLOv5s, and the detection effect of various complex scenes is significantly improved. In addition, the paper analyzes and discusses the generalization ability of YOLO-GBS model on a larger-scale pest data set. This research provides a more accurate and efficient intelligent detection method for rice pests and others crop pests.

Prediction and visualization of gene modulated ultralow cadmium accumulation in brown rice grains by hyperspectral imaging.

Monitoring (including prediction and visualization) the gene modulated cadmium (Cd) accumulation in rice grains is one of the most important steps for identification of key transporter genes responsible for grain Cd accumulation and breeding low grain-Cd-accumulating rice cultivars. A method to predict and visualize the gene modulated ultralow Cd accumulation in brown rice grains based on the hyperspectral image (HSI) technology is proposed in this study. Firstly, the Vis-NIR HSIs of brown rice grain samples with 48Cd content levels induced by gene modulation (ranging from 0.0637 to 0.1845 mg/kg) are collected using HSI system. Then, Kernel-ridge (KRR) and random forest (RFR) regression models based on full spectral data and the data after feature dimension reduction (FDR) with kernel principal component analysis (KPCA) and truncated singular value decomposition (TSVD) algorithms are established to predict the Cd contents. RFR model shows poor performance due to the over-fitting based on the full spectral data, while the KRR model can obtain a good predict accuracy with Rp2 of 0.9035, RMSEP of 0.0037 and RPD of 3.278. After the FDR of the full spectral data, the RFR model combined with TSVD reaches the optimum prediction accuracy with Rp2 of 0.9056, RMSEP of 0.0074 and RPD of 3.318, and the best prediction precision of KRR model can also be further enhanced by TSVD with Rp2 of 0.9224, RMSEP of 0.0067 and RPD of 3.512. Finally, the visualization of the predicted Cd accumulation in brown rice grains are realized based on the best regression model (KRR + TSVD). The results of this work indicate that Vis-NIR HSI has great potential for detection and visualization gene modulation induced ultralow Cd accumulation and transport in rice crops.

An efficient approach to detect and track winter flush growth of litchi tree based on UAV remote sensing and semantic segmentation.

The immature winter flush affects the flower bud differentiation, flowering and fruit of litchi, and then seriously reduces the yield of litchi. However, at present, the area estimation and growth process monitoring of winter flush still rely on manual judgment and operation, so it is impossible to accurately and effectively control flush. An efficient approach is proposed in this paper to detect the litchi flush from the unmanned aerial vehicle (UAV) remoting images of litchi crown and track winter flush growth of litchi tree. The proposed model is constructed based on U-Net network, of which the encoder is replaced by MobeilNetV3 backbone network to reduce model parameters and computation. Moreover, Convolutional Block Attention Module (CBAM) is integrated and convolutional layer is added to enhance feature extraction ability, and transfer learning is adopted to solve the problem of small data volume. As a result, the Mean Pixel Accuracy (MPA) and Mean Intersection over Union (MIoU) on the flush dataset are increased from 90.95% and 83.3% to 93.4% and 85%, respectively. Moreover, the size of the proposed model is reduced by 15% from the original model. In addition, the segmentation model is applied to the tracking of winter flushes on the canopy of litchi trees and investigating the two growth processes of litchi flushes (late-autumn shoots growing into flushes and flushes growing into mature leaves). It is revealed that the growth processes of flushes in a particular branch region can be quantitatively analysed based on the UAV images and the proposed semantic segmentation model. The results also demonstrate that a sudden drop in temperature can promote the rapid transformation of late-autumn shoots into flushes. The method proposed in this paper provide a new technique for accurate management of litchi flush and a possibility for the area estimation and growth process monitoring of winter flush, which can assist in the control operation and yield prediction of litchi orchards.

Optical modeling of organic solar cells based on rubrene and C70.

Optical modeling based on the transfer matrix method is employed to investigate the performance of the organic planar heterojunction solar cell with rubrene/C70 as the active layer. The detailed investigation is directed into the effects of layer thickness of the rubrene and C70 on the total absorbed photon density in the active layer. It is revealed that the optical interference plays important role in the performance of the device and the optimal device performance is achieved when the thicknesses of the rubrene and C70 are set as 33 and 28 nm. The simulated results are also confirmed by the experimental data.

Polysaccharide prediction in Ganoderma lucidum fruiting body by hyperspectral imaging.

Ganoderma lucidum is a traditional Chinese healthy food with many kinds of nutritious activities, and polysaccharide is one of its main active components. Ganoderma lucidum polysaccharide plays a vital role in improving human immunity and anti-oxidation. At present, the methods of detecting polysaccharide content of Ganoderma lucidum are destructive, and the steps are complicated and time-consuming. This study aims to explore the possibility of using hyperspectral imaging (HSI) to predict polysaccharide content in a nondestructive way during the growth of Ganoderma lucidum. The partial least square regression (PLSR) model shows good performance for Ganoderma lucidum ( R p 2  = 0.924, R P D p  = 3.622) with pretreatment method of Savitzky-Golay (SG) and standard normal variate (SNV), and feature selection method of successive projections algorithm (SPA). This study indicates that HSI can quickly and nondestructive detect the polysaccharide content of Ganoderma lucidum, provide guidance for the cultivation industry and improve the economic benefits of Ganoderma lucidum.

Polydopamine/graphene/MnO2 composite-based electrochemical sensor for in situ determination of free tryptophan in plants.

The in vivo detection of small active molecules in plant tissues is essential for the development of precision agriculture. Tryptophan (Trp) is an important precursor material for auxin biosynthesis in plants, and the detection of Trp levels in plants is critical for regulating the plant growth process. In this study, an electrochemical plant sensor was fabricated by electrochemically depositing a polydopamine (PDA)/reduced graphene oxide (RGO)-MnO2 nanocomposite onto a glassy carbon electrode (GCE). PDA/RGO-MnO2/GCE exhibited high electrocatalytic activity for the oxidation of Trp owing to the combined selectivity of PDA and catalytic activity of RGO-MnO2. To address the pH variability of plants, a reliable Trp detection program was proposed for selecting an appropriate quantitative detection model for the pH of the plant or plant tissue of interest. Therefore, a series of linear regression curves was constructed in the pH range of 4.0-7.0 using the PDA/RGO-MnO2/GCE-based sensor. In this pH range, the linear detection range of Trp was 1-300 µM, the sensitivity was 0.39-1.66 µA µM-1, and the detection limit was 0.22-0.39 µM. Moreover, the practical applicability of the PDA/RGO-MnO2/GCE-based sensor was successfully demonstrated by determining Trp in tomato fruit and juice. This sensor stably and reliably detected Trp levels in tomatoes in vitro and in vivo, demonstrating the feasibility of this research strategy for the development of electrochemical sensors for measurements in various plant tissues.

Color and transparency-switchable semitransparent polymer solar cells towards smart windows.

Semitransparent polymer solar cells (ST-PSCs) have attracted worldwide attention owing to unique superiority in multiple utilization of incident light. However, the color of ST-PSCs is relatively uniform after fabrication, cannot be dynamically tuned in terms of application requirement. Herein, we demonstrate a high-efficiency ST-PSCs as a smart window, which can be reversibly switched on and off by a gasochromic tungsten trioxide/platinum (WO3/Pt) back reflector layer. The ST-PSCs can be switchable between colored and bleached states with fast response speed of sub-second during hydrogen exposure. Meanwhile, the color and transparency-switching enable light trapping enhancement in long wavelength range, which can systematically improve power conversion efficiency (PCE). As a result, the ST-PSCs contribute a PCE of 10.2% and 9.1% as well as corresponding average visible transmission (AVT) of 25.4% and 33.8% at colored state and bleached state, respectively, which can meet the visual aesthetics requirement well in building integrated photovoltaics. To the best of our knowledge, this is the first example for ST-PSCs that achieve both color-switching and light trapping. Furthermore, the smart windows facing to automobile sunroof are proposed to prove a practical application towards commercialization. We believe that smart windows with gasochromic functions can promise potential opportunities and directions for the future development of ST-PSCs.

Highly Efficient and High Peak Transmittance Colorful Semitransparent Organic Solar Cells with Hybrid-Electrode-Mirror Microcavity Structure.

Microcavity is an efficient approach to manufacture colorful semitransparent organic solar cells (ST-OSCs) with high color purity by tailoring the transmission spectrum to narrow peaks. However, in this type of colorful semitransparent devices, high power conversion efficiency (PCE) and high peak transmittance are not yet simultaneously achieved. This paper proposes a new type of microcavity structure to achieve colorful ST-OSCs with both high PCE and high peak transmittance, in which a hybrid Au/Ag electrode is used as a mirror and WO3 is used as a spacer layer. First, it is demonstrated that the hybrid Au/Ag electrode mirror brings about an improvement of 7.7 and 5.5% for PCE and peak transmittance, respectively, when compared with those of the reference devices using the Ag electrode mirror. Specifically, the PCE of the optimized devices reaches the satisfactory value of over 9%, and the peak transmittance is over 25%. This value of PCE is the highest one reported so far for the microcavity-based ST-OSCs with the same peak transmittance. Second, it is demonstrated that the second-order resonance of the microcavity can be used to improve the color purity of green ST-OSCs by narrowing the transmission peak, and the combination of the second-order and third-order resonance can be used to construct colorful ST-OSCs with mixed colors. Thus, a novel approach is developed to tune the color of ST-OSCs, which is based on high-order resonance modes of the microcavity.

Decreased Charge Transport Barrier and Recombination of Organic Solar Cells by Constructing Interfacial Nanojunction with Annealing-Free ZnO and Al Layers.

To overcome drawbacks of the electron transport layer, such as complex surface defects and unmatched energy levels, we successfully employed a smart semiconductor-metal interfacial nanojunciton in organic solar cells by evaporating an ultrathin Al interlayer onto annealing-free ZnO electron transport layer, resulting in a high fill factor of 73.68% and power conversion efficiency of 9.81%. The construction of ZnO-Al nanojunction could effectively fill the surface defects of ZnO and reduce its work function because of the electron transfer from Al to ZnO by Fermi level equilibrium. The filling of surface defects decreased the interfacial carrier recombination in midgap trap states. The reduced surface work function of ZnO-Al remodulated the interfacial characteristics between ZnO and [6,6]-phenyl C71-butyric acid methyl ester (PC71BM), decreasing or even eliminating the interfacial barrier against the electron transport, which is beneficial to improve the electron extraction capacity. The filled surface defects and reduced interfacial barrier were realistically observed by photoluminescence measurements of ZnO film and the performance of electron injection devices, respectively. This work provides a simple and effective method to simultaneously solve the problems of surface defects and unmatched energy level for the annealing-free ZnO or other metal oxide semiconductors, paving a way for the future popularization in photovoltaic devices.

Boosted Electron Transport and Enlarged Built-In Potential by Eliminating the Interface Barrier in Organic Solar Cells.

A smart interface modification strategy was employed to simultaneously improve short-circuit current density (Jsc) and open-circuit voltage (Voc) by incorporating a poly[(9,9-bis(3'-(N,N-dimethylamion)propyl)-2,7-fluorene)-alt-2,7-(9,9-dioctyl)-fluorene] (PFN) interlayer between a TiO2 film and an active layer, arising from the fact that PFN effectively eliminated the interface barrier between TiO2 and the fullerene acceptor. The work function (WF) of TiO2 was apparently reduced, which facilitated effective electron transfer from the active layer to the TiO2 electron transport layer (ETL) and suppressed charge carrier recombination between contact interfaces. Electron injection devices with and without a PFN interlayer were fabricated to prove the eliminated electron barrier, meanwhile photoluminescence (PL) and time-resolved transient photoluminescence (TRTPL) were measured to probe much easier electron transfer from [6,6]-phenyl C71-butyric acid methyl ester (PC71BM) acceptor to TiO2 ETL, contributing to enhanced Jsc. The shift in vacuum level altered the WF of PC71BM, which enlarged the internal electrical field at the donor/acceptor interface and built-in potential (Vbi) across the device. Dark current characteristics and Mott-Schottky measurements indicated the enhancement of Vbi, benefiting to increased Voc. Consequently, the champion power conversion efficiency for a device with a PFN interlayer of 0.50 mg/mL reached to 7.14%, which is much higher than the PCE of 5.76% for the control device.

Achieving ultranarrow graphene perfect absorbers by exciting guided-mode resonance of one-dimensional photonic crystals.

Graphene perfect absorbers with ultranarrow bandwidth are numerically proposed by employing a subwavelength dielectric grating to excite the guided-mode resonance of one-dimensional photonic crystals (1DPCs). Critical coupling of the guided-mode resonance of 1DPCs to graphene can produce perfect absorption with a ultranarrow bandwidth of 0.03 nm. The quality factor of the absorption peak reaches a ultrahigh value of 20000. It is also found that the resonant absorption peaks can be tuned by controlling the dispersion line of the guided mode and the period of the grating. When the parameters of the grating and the 1DPCs are suitably set, the perfect absorption peaks can be tuned to any randomly chosen wavelength in the visible wavelength range.

Highly efficient semitransparent polymer solar cells with color rendering index approaching 100 using one-dimensional photonic crystal.

Window application is the important aim for semitransparent solar cells (STPSC) investigation. Here, we demonstrate a method to achieve significantly improved color rendering index (CRI), depressed chromaticity difference (DC), and enhanced power conversion efficiency (PCE) simultaneously by introducing the one-dimensional photonic crystals (1DPCs) Bragg reflector structure onto the STPSC. The device performance is studied from aspects of color perception, electrical properties, and theoretical optical simulations. The STPSCs exhibit achromatic transparency nature color perceptions, especially for the STPSCs with 1DPCs (pairs ≥ 3) under AM 1.5G illumination light source, standard illuminant D65, and standard illuminant A. The excellent CRI is approaching 97 with lower DC about 0.0013 for the device with 5 pairs of 1DPC illumined by AM 1.5G illumination light source. At the same time, the PCE of STPSC devices with 5 pairs of 1DPC was improved from 4.87 ± 0.14% to 5.31 ± 0.13% compared to without. This method provides a facilitative approach to realizing excellent SPTSC window application.