Multispectral Sources Research Articles

Cross-directional consistency network with adaptive layer normalization for multi-spectral vehicle re-identification and a high-quality benchmark

To tackle the challenge of vehicle re-identification (Re-ID) in complex lighting environments and diverse scenes, multi-spectral sources like visible and infrared information are taken into consideration due to their excellent complementary advantages. However, multi-spectral vehicle Re-ID suffers cross-modality discrepancy caused by heterogeneous properties of different modalities as well as a big challenge of the diverse appearance with different views in each identity. Meanwhile, diverse environmental interference leads to heavy sample distributional discrepancy in each modality. In this work, we propose a novel cross-directional consistency network (CCNet) to simultaneously overcome the discrepancies from both modality and sample aspects. In particular, we design a new cross-directional center loss (Lcdc) to pull the modality centers of each identity close to mitigate cross-modality discrepancy, while the sample centers of each identity close to alleviate the sample discrepancy. Such a strategy can generate discriminative multi-spectral feature representations for vehicle Re-ID. In addition, we design an adaptive layer normalization unit (ALNU) to dynamically adjust individual feature distribution to handle distributional discrepancy of intra-modality features for robust learning. To provide a comprehensive evaluation platform, we create a high-quality RGB-NIR-TIR multi-spectral vehicle Re-ID benchmark (MSVR310), including 310 different vehicles from a broad range of viewpoints, time spans and environmental complexities. Comprehensive experiments on both created and public datasets demonstrate the effectiveness of the proposed approach comparing to the state-of-the-art methods. The dataset and code will be released for free academic usage at https://github.com/superlollipop123/Cross-directional-Center-Network-and-MSVR310.

Nonlinear frequency conversion of a Ho :YAG laser beam

We report the results of experiments on frequency conversion of a Ho : YAG laser beam (wavelength λ ∼ 2.1 μm) into the near-, mid-, and far-IR regions, aimed at expanding the spectral composition of compact multispectral sources of coherent radiation. The experimentally found laser conversion efficiency into the second harmonic reaches 32 % and 54 % in the cw and repetitively pulsed regimes, respectively. The parametric conversion efficiency into the mid-IR range (λ = 3.5 – 5.0 μm) reaches 55 %, while for the far-IR range (λ = 8 – 8.2 μm) it turns out to be 10.5 %.

Ultra-broadband contactless imaging power meter.

Knowledge of the spatial and temporal distribution of heat flux is of great interest for the quantification of heat sources. In this work, we describe the development of a new ultra-broadband contactless imaging power meter based on electromagnetic to infrared technology. This new sensor and the mathematical processing of images enable the reconstruction of both spatial and amplitude distributions through a wide spectral range of sources. The full modeling of the thermoconverter based on 3D formalism of thermal quadrupoles is presented first before deriving a reduced model more suitable for quick and robust inverse processing. The inverse method makes it possible to simultaneously identify the heat losses and the spatial and temporal source distribution for the first time, to the best of our knowledge. Finally, measurements of multispectral sources are presented and discussed, with an emphasis on the spatial and temporal resolution, accuracy and capabilities of the power meter.

Fiber-coupling optical system for high-power and multi-wavelength diode laser bars oriented to integrated biomedical imaging systems

The use of pulsed high-power diode lasers (HPDLs) in the near infrared (NIR) range as multispectral sources has attracted much interest in biomedical applications due to their relatively low cost and small size compared to nanosecond Nd:YAG optical parametric oscillator (OPO) lasers. One of the main limitations of these sources is the availability to combine different wavelengths with high power density in the same beam. Various works have shown that the use of linear arrays of emitters as diode laser bars (DLBs) or stacked arrays of emitters as diode laser stacks (DLSs) allows the combination of multiple wavelengths while maintaining high power densities. Nevertheless, the highly asymmetric beam profile emitted by such laser sources between fast and slow axes implies the need for suitable beam shaping for efficient fiber coupling. In this work, we investigate a novel beam shaping technique to homogenize the beam quality of six DLBs in the wavelength range between 790 nm and 980 nm.We consider fast-axis collimating lenses (FAC) and beam twisters to reduce the beam asymmetry of the individual bars. The beams from the DLBs are then combined into a single multispectral beam using reflective mirrors, dichroic mirrors, and a polarizing beam splitter cube (PBC), and effectively coupled into a 400 µm core-diameter/N.A. = 0.22 optical fiber using a pair of cylindrical lenses. Simulation shows high coupled power densities with ~ 1.8 MW/cm2 at the output of the fiber. The coupling efficiency reaches 89.4%. The use of submillimeter fiber optic probes is particularly promising for photoacoustic endoscopy (PAE) applications requiring minimally invasive examination of internal organs.

Attention fusion network for multi-spectral semantic segmentation

To improve the accuracy of multi-spectral semantic segmentation, an attention fusion network (AFNet) based on deep learning is proposed. Different from current methods, the AFNet uses a co-attention mechanism by designing an attention fusion module to calculate the spatial correlation between the red-green-blue (RGB) image and infrared (IR) image feature maps to guide the fusion of features from different spectra. This approach enhances the feature presentation and makes full use of the complementary characteristics of multi-spectral sources. The proposed network is tested on RGB-IR datasets and compared with relevant state-of-the-art networks. The experimental analyses prove that the proposed AFNet can improve multi-spectral semantic segmentation results with good visual definition and high accuracy in classification and localization.

Silicon arrayed waveguide gratings at 20-μm wavelength characterized with an on-chip resonator

Low-loss arrayed waveguide gratings (AWGs) are demonstrated at a 2.0-μm wavelength. These devices promote rapidly developing photonic applications, supported by the recent development of mid-infrared lasers integrated on silicon (Si). Multi-spectral photonic integrated circuits at 2.0-μm are envisioned since the AWGs are fabricated with the 500-nm-thick Si-on-insulator platform compatible with recently demonstrated lasers and semiconductor optical amplifiers on Si. Characterization with the AWG-ring method improves the on-chip transmission uncertainty to ∼6% compared to the conventional method with an uncertainty of ∼53%. Channel losses of ∼2.4 dB are found, with -31 dB crosstalk per channel. Fully integrated multi-spectral sources at 2.0μm with pump lasers, low-loss multiplexers, and an output amplifier are now feasible.

A multispectral linear optical sweep source, based on tunable InGaN/GaN QW LEDs

Applications of multispectral sources are numerous. The previous forms of optical sources are continuously being replaced by efficient LEDs with an emphasis on III-nitrides, since it covers a very wide optical range. In this paper, we are proposing a multispectral linear optical sweep source, based on tunable InGaN/GaN QW LEDs. For linearization, a λ verses current curve of tunable InGaN/GaN QW LED, as obtained from our well established method of solution of Schrödinger and Poisson equations is considered as a model. The entire process of linearization is outlined in a simple form, for clear understanding. The linear sweep is so designed that it can be stopped at any λ,. i.e. color. Changing of the range of the spectral sweep through changes in the device parameters is pointed out. Procedures, necessary constituents and results are presented with suitable discussions. The multispectral ramp should be of much use for a wide range of applications.

CMOS compatible metal-insulator-metal plasmonic perfect absorbers

We report the design, fabrication and characterization of CMOS compatible metal-insulator-metal (MIM) plasmonic resonators made of tungsten and silicon nitride for the mid-infrared range. These structures give rise to spectrally selective emission/absorption, which is of particular interest in the field of non-dispersive infrared (NDIR) gas spectroscopy. In this paper, we demonstrate large scale fabrication on 200 mm silicon wafer of such devices, and show some of their main characteristics, such as tunability, multi-spectral sources, polarization-independance and consistency between reflectivity and emissivity measurements.

Microscopía holográfica digital sin lentes con resolución micrométrica y fuentes multiespectrales

Digital lensless holographic microscopy (DLHM) as an imaging architecture with micrometer spatial resolution and operating with multispectral sources is shown. The principles and recent advances in the realm of this technology of microscopy are presented from the point of view of the contributions of our group. DLHM operating a full color and with femto-second lasers are the advances highlighted in this work. The effects of the spatial and the temporal coherence of the light source on the performance of the DLHM are analyzed by comparing the reconstructed images of the microscope by using a LED and a 12 fs mode-locked Ti:Sa laser. A section of the head of a Drosophila melanogaster fly is utilized as a sample with complex internal structure. © 2015. Acad. Colomb. Cienc. Ex. Fis. Nat. All rights reserved.

Evaluating leaf chlorophyll content prediction from multispectral remote sensing data within a physically-based modelling framework

Accurate modelling of leaf chlorophyll content over a range of spatial and temporal scales is central to monitoring vegetation stress and physiological condition, and vegetation response to different ecological, climatic and anthropogenic drivers. A process-based modelling approach can account for variation in other factors affecting canopy reflectance, providing a more accurate estimate of chlorophyll content across different vegetation species, time-frames, and broader spatial extents. However, physically-based modelling studies usually use hyperspectral data, neglecting a wealth of data from broadband and multispectral sources. In this study, we assessed the potential for using canopy (4-Scale) and leaf radiative transfer (PROSPECT4/5) models to estimate leaf chlorophyll content using canopy Landsat satellite data and simulated Landsat bands from leaf level hyperspectral reflectance data. Over 600 leaf samples were used to test the performance of PROSPECT for different vegetation species, including black spruce (Picea mariana), sugar maple (Acer saccharum), trembling aspen (Populus tremuloides) and jack pine (Pinus banksiana). At the leaf level, hyperspectral and simulated Landsat bands showed very similar results to laboratory measured chlorophyll (R2=0.77 and R2=0.75, respectively). Comparisons between PROSPECT4 modelled chlorophyll from simulated Landsat and hyperspectral spectra showed a very close correspondence (R2=0.97, root mean square error (RMSE)=3.01μg/cm2), as did simulated reflectance bands from other broadband and narrowband sensors (MODIS: R2=0.99, RMSE=1.80μg/cm2; MERIS: R2=0.97, RMSE=2.50μg/cm2 and SPOT5 HRG: R2=0.96, RMSE=5.38μg/cm2). Modelled leaf chlorophyll content from Landsat 5 TM canopy reflectance data, acquired from over 40 ground validation sites, demonstrated a strong relationship with measured leaf chlorophyll content (R2=0.78, RMSE=8.73μg/cm2, p<0.001), and a high linearity with negligible systematic bias. Study results demonstrate the small number of input bands required for PROSPECT inversion and provide a theoretical and operational basis for the future retrieval of leaf chlorophyll content using broadband or multispectral sensors within a physically-based approach.