Digital Grading the Color Fastness to Rubbing of Fabrics Based on Spectral Reconstruction and BP Neural Network.

To digital grade the staining color fastness of fabrics after rubbing, an automatic grading method based on spectral reconstruction technology and BP neural network was proposed. Firstly, the modeling samples are prepared by rubbing the fabrics according to the ISO standard of 105-X12. Then, to comply with visual rating standards for color fastness, the modeling samples are professionally graded to obtain the visual rating result. After that, a digital camera is used to capture digital images of the modeling samples inside a closed and uniform lighting box, and the color data values of the modeling samples are obtained through spectral reconstruction technology. Finally, the color fastness prediction model for rubbing was constructed using the modeling samples data and BP neural network. The color fastness level of the testing samples was predicted using the prediction model, and the prediction results were compared with the existing color difference conversion method and gray scale difference method based on the five-fold cross-validation strategy. Experiments show that the prediction model of fabric color fastness can be better constructed using the BP neural network. The overall performance of the method is better than the color difference conversion method and the gray scale difference method. It can be seen that the digital rating method of fabric staining color fastness to rubbing based on spectral reconstruction and BP neural network has high consistency with the visual evaluation, which will help for the automatic color fastness grading.

Super tiny quartz-tuning-fork-based light-induced thermoelastic spectroscopy sensing.

In this Letter, a sensitive light-induced thermoelastic spectroscopy (LITES)-based trace gas sensor by exploiting a super tiny quartz tuning fork (QTF) was demonstrated. The prong length and width of this QTF are 3500 µm and 90 µm, respectively, which determines a resonant frequency of 6.5 kHz. The low resonant frequency is beneficial to increase the energy accumulation time in a LITES sensor. The geometric dimension of QTF on the micrometer scale is advantageous to obtain a great thermal expansion and thus can produce a strong piezoelectric signal. The temperature gradient distribution of the super tiny QTF was simulated based on the finite element analysis and is higher than that of the commercial QTF with 32.768 kHz. Acetylene (C2H2) was used as the analyte. Under the same conditions, the use of the super tiny QTF achieved a 1.64-times signal improvement compared with the commercial QTF. The system shows excellent long-term stability according to the Allan deviation analysis, and a minimum detection limit (MDL) would reach 190 ppb with an integration time of 220 s.

Research on the deep learning-based exposure invariant spectral reconstruction method.

The surface spectral reflectance of an object is the key factor for high-fidelity color reproduction and material analysis, and spectral acquisition is the basis of its applications. Based on the theoretical imaging model of a digital camera, the spectral reflectance of any pixels in the image can be obtained through spectral reconstruction technology. This technology can avoid the application limitations of spectral cameras in open scenarios and obtain high spatial resolution multispectral images. However, the current spectral reconstruction algorithms are sensitive to the exposure variant of the test images. That is, when the exposure of the test image is different from that of the training image, the reconstructed spectral curve of the test object will deviate from the real spectral to varying degrees, which will lead to the spectral data of the target object being accurately reconstructed. This article proposes an optimized method for spectral reconstruction based on data augmentation and attention mechanisms using the current deep learning-based spectral reconstruction framework. The proposed method is exposure invariant and will adapt to the open environment in which the light is easily changed and the illumination is non-uniform. Thus, the robustness and reconstruction accuracy of the spectral reconstruction model in practical applications are improved. The experiments show that the proposed method can accurately reconstruct the shape of the spectral reflectance curve of the test object under different test exposure levels. And the spectral reconstruction error of our method at different exposure levels is significantly lower than that of the existing methods, which verifies the proposed method's effectiveness and superiority.

An improved spectral estimation method based on color perception features of mobile phone camera.

We use the mobile phone camera as a new spectral imaging device to obtain raw responses of samples for spectral estimation and propose an improved sequential adaptive weighted spectral estimation method. First, we verify the linearity of the raw response of the cell phone camera and investigate its feasibility for spectral estimation experiments. Then, we propose a sequential adaptive spectral estimation method based on the CIE1976 L*a*b* (CIELAB) uniform color space color perception feature. The first stage of the method is to weight the training samples and perform the first spectral reflectance estimation by considering the Lab color space color perception features differences between samples, and the second stage is to adaptively select the locally optimal training samples and weight them by the first estimated root mean square error (RMSE), and perform the second spectral reconstruction. The novelty of the method is to weight the samples by using the sample in CIELAB uniform color space perception features to more accurately characterize the color difference. By comparing with several existing methods, the results show that the method has the best performance in both spectral error and chromaticity error. Finally, we apply this weighting strategy based on the CIELAB color space color perception feature to the existing method, and the spectral estimation performance is greatly improved compared with that before the application, which proves the effectiveness of this weighting method.

Highly sensitive HF detection based on absorption enhanced light-induced thermoelastic spectroscopy with a quartz tuning fork of receive and shallow neural network fitting.

Due to its advantages of non-contact measurement and high sensitivity, light-induced thermoelastic spectroscopy (LITES) is one of the most promising methods for corrosive gas detection. In this manuscript, a highly sensitive hydrogen fluoride (HF) sensor based on LITES technique is reported for the first time. With simple structure and strong robustness, a shallow neural network (SNN) fitting algorithm is introduced into the field of spectroscopy data processing to achieve denoising. This algorithm provides an end-to-end approach that takes in the raw input data without any pre-processing and extracts features automatically. A continuous wave (CW) distributed feedback diode (DFB) laser with an emission wavelength of 1.27 µm was used as the excitation source. A Herriott multi-pass cell (MPC) with an optical length of 10.1 m was selected to enhance the laser absorption. A quartz tuning fork (QTF) with resonance frequency of 32,767.52 Hz was adopted as the thermoelastic detector. An Allan variance analysis was performed to demonstrate the system stability. When the integration time was 110 s, the minimum detection limit (MDL) was found to be 71 ppb. After the SNN fitting algorithm was used, the signal-to-noise ratio (SNR) of the HF-LITES sensor was improved by a factor of 2.0, which verified the effectiveness of this fitting algorithm for spectroscopy data processing.

Optimized method for segmentation of ancient mural images based on superpixel algorithm.

High-precision segmentation of ancient mural images is the foundation of their digital virtual restoration. However, the complexity of the color appearance of ancient murals makes it difficult to achieve high-precision segmentation when using traditional algorithms directly. To address the current challenges in ancient mural image segmentation, an optimized method based on a superpixel algorithm is proposed in this study. First, the simple linear iterative clustering (SLIC) algorithm is applied to the input mural images to obtain superpixels. Then, the density-based spatial clustering of applications with noise (DBSCAN) algorithm is used to cluster the superpixels to obtain the initial clustered images. Subsequently, a series of optimized strategies, including (1) merging the small noise superpixels, (2) segmenting and merging the large noise superpixels, (3) merging initial clusters based on color similarity and positional adjacency to obtain the merged regions, and (4) segmenting and merging the color-mixing noisy superpixels in each of the merged regions, are applied to the initial cluster images sequentially. Finally, the optimized segmentation results are obtained. The proposed method is tested and compared with existing methods based on simulated and real mural images. The results show that the proposed method is effective and outperforms the existing methods.

An integrated method for color correction based on color constancy for early mural images in Mogao Grottoes.

Restoring the correct or realistic color of a cultural heritage object is a crucial problem for imaging techniques. Digital images often have undesired color casts due to adverse effects caused by unstable illuminant conditions, vignetting, and color changes due to camera settings. In this work, we present an improved color correction method for color cast images that makes the color appear more realistic. It is based on a computational model of the human visual system that perceives objects by color constancy theory; it realizes illumination non-uniformity compensation and chromaticity correction for color cast images by taking into account the color stability of some pigments. This approach has been used to correct the color in Cave 465 of the Mogao Grottoes. The experimental results demonstrate that the proposed method is able to "adaptively correct" color cast images with widely varying lighting conditions and improve the consistency efficaciously. It can achieve improved consistency in the mean CIEDE2000 color difference compared with the images before correction. This colorimetric correction methodology is sufficiently accurate in color correction implementation for cast images of murals captured in the early years.

Towards an optimum colour preference metric for white light sources: a comprehensive investigation based on empirical data.

Colour preference is a critical dimension for describing the colour quality of lighting and numerous metrics have been proposed. However, due to the variation amongst psychophysical studies, consensus has not been reached on the best approach to quantify colour preference. In this study, 25 typical colour quality metrics were comprehensively tested based on 39 groups of psychophysical data from 19 published visual studies. The experimental results showed that two combined metrics: the arithmetic mean of the gamut area index (GAI) and colour rendering index (CRI) and the colour quality index (CQI), a combination of the correlated colour temperature (CCT) and memory colour rendering index (MCRI), exhibit the best performance. Qp in the colour quality scale (CQS) and MCRI also performed well in visual experiments of constant CCT but failed when CCT varied, which highlights the dependence of certain metrics on contextual lighting conditions. In addition, it was found that some weighted combinations of an absolute gamut-based metric and a colour fidelity metric exhibited superior performance in colour preference prediction. Consistent with such a result, a novel metric named MCPI (colour preference index based on meta-analysis) was proposed by fitting the large psychophysical dataset, and this achieved a significantly higher weighted average correlation coefficient between metric predictions and subjective preference ratings.

Sequential adaptive estimation for spectral reflectance based on camera responses.

A sequential weighted nonlinear regression technique from digital camera responses is proposed for spectral reflectance estimation. The method consists of two stages taking colorimetric and spectral errors between training set and target set into accounts successively. Based on polynomial expansion model, local optimal training samples are adaptively employed to recover spectral reflectance as accurately as possible. The performance of the method is compared with several existing methods in the cases of simulated camera responses under three kinds of noise levels and practical camera responses under the self as well as cross test conditions. Results show that the proposed method is able to recover spectral reflectance with a higher accuracy than other methods considered.

Spectra estimation from raw camera responses based on adaptive local-weighted linear regression.

An improved spectral reflectance estimation method is developed to transform raw camera RGB responses to spectral reflectance. The novelty of our method is to apply a local weighted linear regression model for spectral reflectance estimation and construct the weighting matrix using a Gaussian function in CIELAB uniform color space. The proposed method was tested using both a standard color chart and a set of textile samples, with a digital RGB camera and by ten times ten-fold cross-validation. The results demonstrate that our method gives the best accuracy in estimating both the spectral reflectance and the colorimetric values in comparison with existing methods.

Study on Dual Channel Lateral Field Excitation Quartz Crystal Microbalance for Measuring Liquid Electrical Properties.

Lateral field excitation quartz crystal microbalance (LFE-QCM) can detect both the electrical properties (conductivity and permittivity) and mechanical properties (viscosity and density) of the liquid. In practical applications for detecting electrical properties, the viscosity and density of the liquid will also change. This research proposed a dual-channel LFE-QCM for reducing the influence of density and viscosity. The sensing layer of one resonant element is almost bare, and the other is covered by a metal film as a reference. Different organic solutions and NaCl solution were used to study the influence of mechanical properties and the temperature on electrical properties. The experimental results demonstrate that the dual-channel LFE-QCM is necessary for properly detecting electrical properties of the liquid.

Development of a Flow Injection Based High Frequency Dual Channel Quartz Crystal Microbalance.

When the quartz crystal microbalance (QCM) is used in liquid for adsorption or desorption monitoring based bio- or chemical sensing applications, the frequency shift is not only determined by the surface mass change, but also by the change of liquid characteristics, such as density and viscosity, which are greatly affected by the liquid environmental temperature. A monolithic dual-channel QCM is designed and fabricated by arranging two QCM resonators on one single chip for cancelling the fluctuation induced by environmental factors. In actual applications, one QCM works as a specific sensor by modifying with functional membranes and the other acts as a reference, only measuring the liquid property. The dual-channel QCM is designed with an inverted-mesa structure, aiming to realize a high frequency miniaturized chip and suppress the frequency interference between the neighbored QCM resonators. The key problem of dual-channel QCMs is the interference between two channels, which is influenced by the distance of adjacent resonators. The diameter of the reference electrode has been designed into several values in order to find the optimal parameter. Experimental results demonstrated that the two QCMs could vibrate individually and the output frequency stability and drift can be greatly improved with the aid of the reference QCM.

[Research on Visual Perception-Referenced Compression Method for Multi-Spectral Data with High-Fidelity].

Aim In order to maintain the chromaticity precision in the process of linear compression of the multispectral data, a visual perception-referenced compression method (VPCM) based on the chroma gradient (refer to the partial derivative of chroma to wavelength) is proposed. Method The method firstly successfully developed the transfer functions which could synchronously fusion the spectral features and chromaticity characteristics of human visuals based on the nonlinear analytic feature of human visual system. For further improvement the transfer function, a modified optimizing function was developed to help find out the optimal transfer direction for different sample sets. If the transfer function was finally settled, it will be applied to transforming the spectral data of the sample set (Γ(S)=C). Then the transformed spectral data of the sample set will be compressed with high chromatic accuracy by the principle components analysis method. After that, the compressed data will be reconstructed through inverse transformation (Γ(-1)(C)=), while the reconstructed spectral data will be using to evaluate the effective of the proposed VPCM method. Result Four groups typical and representative sample sets were chosen to test the effective of the proposed method. The CIELab color difference in the D50/2° calculates condition and a proposed mean metamerism index (MMI) calculated with 75 groups typical light sources (including tungsten, fluorescent and LED lamp) was adopted as evaluating metrics. Eventually, the comparative experiment involving several existing methods Lab-PQR and 2-XYZ indicates that the proposed VPCM hold the best chromatic accuracy both for metric MMI and the average color difference ΔE(ab) when compared with Lab-PQR and 2-XYZ, and the spectral accuracy was calculated between Lab-PQR and 2-XYZ with Lab-PQR maintained the highest spectral accuracy. Conclusion The proposed VPCM can preserve high compression chromatic precision at the price of small loss of spectral precision and possess good colorimetric stability under variable reference conditions. It is very applicable for some application fields which require compressing of the multi-spectral data with high chromatic accuracy.

In Situ Identification of Pigment Composition and Particle Size on Wall Paintings Using Visible Spectroscopy as a Noninvasive Measurement Method.

Noninvasive examination methods of chemical composition and particle size are presented here based on visible spectroscopy to achieve the identification and recording of mineral pigments used on ancient wall paintings. The normalized spectral curve, slope and curvature extracted from visible spectral reflectance are combined with adjustable weighting coefficients to construct the identification feature space, and Euclid distances between spectral reflectance from wall paintings and a reference database are calculated in the feature space as the discriminant criterion to identify the chemical composition of mineral pigments. A parametric relationship between the integral quantity of spectral reflectance and logarithm of mean particle size is established using a quadratic polynomial to accomplish the noninvasive prediction of mineral pigment particle size used on ancient wall paintings. The feasibility of the proposed methods is validated by the in situ nondestructive identification of the wall paintings in the Mogao Grottoes at Dunhuang. Chinese painting styles and historical evolution are then analyzed according to the identification results of 16 different grottoes constructed from the Sixteen Kingdoms to the Yuan Dynasty.

[Research on the Training Samples Selection for Spectral Reflectance Reconstruction Based on Principal Component Analysis].

The composition of training samples set is an important influence factor of spectral reflectance reconstruction process. Representative color samples selection for learning-based spectral reflectance reconstruction is discussed in this paper. A method based on Principal Component Analysis (PCA) is proposed to perform sample selection. First of all, a part of samples are selected according to the minimum Euclidean distance criteria in terms of camera response value from a large number of samples, which aim to ensure the similarity between training samples and target samples. Then the PCA data processing method is applied to these samples after removing the duplicate samples. The samples with larger principal component loadings are regarded as the representative color samples. Different thresholds for each principal component are used to make decision whether the loading of sample is large enough. In order to validate the proposed method, the selected samples are used as training samples to recover the spectral reflectance of color patches. A real multi-channel imaging system by loading broadband color filters in front of lens is used in the experiment to acquire the multi-channel image dataset. In this paper the pseudo-inverse method is employed to reconstruct spectral reflectance of target color patches. It is shown that the proposed method is superior to the previous methods in spectral reconstruction accuracy and can meet the requirements of high precision color reproduction.

[Research on Developping the Colorimetric Correction Method Based on Wide Gamut Spectral Color Dataset].

In cross-media digital color imaging workflow, the colorimetric information transform among various digital medias always suffers from large transform errors, due to the difference among the lighting illuminants and the digital devices. Colorimetric correction aims at avoiding the colorimetric transform mismatch and thus improving the color transform accuracy comprehensively. Till now, two kinds of colorimetric transform methods have been proposed, which are the chromatic adaptation transform and the regression-based transform. However, since the color gamut of the training samples of such two method are both small, adopting those methods in colorimetric domain always leads to large colorimetric transform errors for the high saturated color regions. In this research, in order to reduce the large correction error in high saturated color regions, a modified colorimetric correction method basing on a wide gamut spectral dataset was proposed. The wide gamut spectral dataset was built by comprehensively collecting and producing typical spectral color samples and could provide optimal training samples for the existing regression based colorimetric correction model, with the help of gamut partition and optimal color purity choosing. By modifying the existing method with such samples, the colorimetric correction performance obviously improved. The experimental result shows that the modified colorimetric correction method performs significantly better than the existing methods and the colorimetric correction errors are successfully reduced by around 15% according to proposed method in form of CIEDE2000 color difference, while as for the high saturated color regions the reduction rate of the colorimetric correction errors approximately grows to 40%. The authors believe that the proposed method will provided effective support for the development of digital color imaging in near future.

Research on the Development of Light Blending Model for Smart LED Lighting.

The color of the LED smart light is tunable by its inner equipped micro-processing systems. Therefore, it could provide significant improvement for the smart lighting conditions, such as museum lighting and home lighting. At present, the limitation of the current lighting blending technology remarkably affects the application of smart lighting technology and people could not make full use of the adjustability of the smart luminaries. In this research, a novel light blending model was proposed based on BP neural network and active set algorithm. The models could effectively simulate the nonlinear relationship between the device control values of the smart light and the output radiance spectrum of the light. Particularly, a BP neural network-based forward model for LED light blending was firstly proposed, which could accurately calculate the spectral radiance power distribution from the device control values. Afterwards, based on forward model, an active set algorithm-based backward model was developed, which could precisely predict the device control values from the desired spectral radiance power distribution. The experimental result indicates that the proposed method could accurately achieve the light blending controlling of smart LED light, with a CIEUCS Duv value of 0.002 7, which is significantly smaller than the just noticeable difference value of human vision. The authors believe that the proposed method will provided effective support for the development of smart LED lighting in near future.

Flip Chip Bonding of a Quartz MEMS-Based Vibrating Beam Accelerometer.

In this study, a novel method to assemble a micro-accelerometer by a flip chip bonding technique is proposed and demonstrated. Both the main two parts of the accelerometer, a double-ended tuning fork and a base-proof mass structure, are fabricated using a quartz wet etching process on Z cut quartz wafers with a thickness of 100 µm and 300 µm, respectively. The finite element method is used to simulate the vibration mode and optimize the sensing element structure. Taking advantage of self-alignment function of the flip chip bonding process, the two parts were precisely bonded at the desired joint position via AuSn solder. Experimental demonstrations were performed on a maximum scale of 4 × 8 mm² chip, and high sensitivity up to 9.55 Hz/g with a DETF resonator and a Q value of 5000 in air was achieved.

Design and implementation of a micromechanical silicon resonant accelerometer.

The micromechanical silicon resonant accelerometer has attracted considerable attention in the research and development of high-precision MEMS accelerometers because of its output of quasi-digital signals, high sensitivity, high resolution, wide dynamic range, anti-interference capacity and good stability. Because of the mismatching thermal expansion coefficients of silicon and glass, the micromechanical silicon resonant accelerometer based on the Silicon on Glass (SOG) technique is deeply affected by the temperature during the fabrication, packaging and use processes. The thermal stress caused by temperature changes directly affects the frequency output of the accelerometer. Based on the working principle of the micromechanical resonant accelerometer, a special accelerometer structure that reduces the temperature influence on the accelerometer is designed. The accelerometer can greatly reduce the thermal stress caused by high temperatures in the process of fabrication and packaging. Currently, the closed-loop drive circuit is devised based on a phase-locked loop. The unloaded resonant frequencies of the prototype of the micromechanical silicon resonant accelerometer are approximately 31.4 kHz and 31.5 kHz. The scale factor is 66.24003 Hz/g. The scale factor stability is 14.886 ppm, the scale factor repeatability is 23 ppm, the bias stability is 23 µg, the bias repeatability is 170 µg, and the bias temperature coefficient is 0.0734 Hz/°C.

An experimental study on fabricating an inverted mesa-type quartz crystal resonator using a cheap wet etching process.

In this study, a miniaturized high fundamental frequency quartz crystal microbalance (QCM) is fabricated for sensor applications using a wet etching technique. The vibration area is reduced in the fabrication of the high frequency QCM with an inverted mesa structure. To reduce the complexity of the side wall profile that results from anisotropic quartz etching, a rectangular vibration area is used instead of the conventional circular structure. QCMs with high Q values exceeding 25,000 at 47 MHz, 27,000 at 60 MHz, 24,000 at 73 MHz and 25,000 at 84 MHz are fabricated on 4 × 4 mm2 chips with small vibration areas of 1.2 × 1.4 mm2. A PMMA-based flow cell is designed and manufactured to characterize the behavior of the fabricated QCM chip in a liquid. Q values as high as 1,006 at 47 MHz, 904 at 62 MHz, 867 at 71 MHz and 747 at 84 MHz are obtained when one side of the chip is exposed to pure water. These results show that fabricated QCM chips can be used for bio- and chemical sensor applications in liquids.