Search

To enhance the heat and mass transfer during dimethyl ether (DME) steam reforming, a micro-reactor with catalyst coated on nickel foam support was designed and fabricated. A two-dimensional numerical model with SIMPLE algorithm and finite volume method was used to investigate 1) the fluid flow, 2) the heat transfer and 3) chemical reactions consist of DME hydrolysis, methanol steam reforming, methanol decomposition and water gas shift reactions. Both the numerical and the experimental results showed that the DME conversion in the micro-reactor is higher than that in the fixed bed reactor. The numerical study also showed that the velocity and the temperature distribution were more uniform in the micro-reactor. Wall temperature, porosity and steam/DME ratio have been investigated in order to optimize the process in the micro-reactor. The wall temperature of 270 °C and the steam/DME feed ratio of 5 were recommended. Meanwhile the results indicate that a larger porosity will give a higher DME conversion and CO concentration.

Preferential oxidation of CO (CO-PrOx) is an important step to meet the need of the proton exchange membrane (PEM) fuel cell without the Pt anion poison. A glycine-nitrate approach was used for the synthesis of Co/CeO2 nanoparticle for preferential oxidation of CO, which a precursor solution was prepared by mixing glycine with an aqueous solution of blended nitrate in stoichiometric ratio. Then the glycine-mixed precursor solution was heated in a beaker for producing nanosized porous powders. Catalytic properties of the powders were investigated and results illustrate that the Co-loading of 30 wt.% catalysts exhibits excellent catalytic properties. Various characterization techniques like X-ray diffraction, SEM, BET, Raman and TPR were used to analyze the relationship between catalyst nature and catalytic performance. The X-ray diffraction patterns and SEM micrographs indicate that catalysts prepared by glycine-nitrate combustion own mesopore structure. The BET, Raman and TPR results showed that the high activity of the 30 wt.% Co-loading of Co/CeO2 catalysts is related to the high BET surface and the strongly interaction between fine-dispersed Co species and CeO2 support.

A bi-function catalyst containing CuZnAlCr and HZSM-5 was used to generate hydrogen by stream reforming of dimethyl ether (SRD) in a metal foam micro-reactor and a fix-bed reactor. Dimethyl ether conversion of 99% and hydrogen yield of >95% was reached with HZSM-5/CuZnAlCr (mass ratio of 1:1) in the micro-reactor. A suitable balance between the dimethyl ether hydrolysis and methanol reforming steps requires the proper acidity and the metal sites. The CuZnAlCr/HZSM-5 properties, effect of CuZnAlCr to HZSM-5 mass ratio were investigated in the metal foam micro-reactor. Moreover, CO was removed from hydrogen-rich gas by preferential oxidation reaction (CO-PrOx) with PtFe/γ-Al 2 O 3 catalyst in a similar metal foam micro-reactor follows the SRD stage. With the optimized O 2 /CO ratio and reaction temperature, the CO concentration dropped to

Reducing the cost of PEM water electrolyser is still a big challenge for renewable energy storage. Though there are some reports which have achieved very low Ir loading (0.1~0.2mg/cm2) in membrane electrode assemblies (MEA) for PEM water electrolyser, the other issues such as the performance and stability of MEA, the reduction of Pt loading in cathode and the scalability of preparation method of MEA still need further investigation for developing low cost and high performance PEM water electrolyser. In this work, to develop high performance anode with low Ir loading, highly active Iridium oxohydroxides were supported on conductive titanium metal nanosphere as the anode catalyst. The diatmeter of metal Ti nanosphere ranges from 60-200nm, and the Iridium oxohydroxides loaded on Ti support shows a particle size between 1.5-3nm. The OER activity of the catalyst with various Ir loading was characterized, and the catalyst with 10% Ir loading exhibited highest Ir mass activity (1700A/g @η=0.37V), it is about 6 times as the commercial IrO2. Increasing the loading of Ir will lead to the decline of Ir mass activity, this may be due to the limited surface area of Ti nanosphere. The stability of the catalyst (Ir loading 10%) was characterized by CV scans between 0.4-1.5V (vs.RHE) in 0.5M H2SO4. After 2000 cycles, the active sites of IrOx declined only 18%, and the catalyst does not show an obviously rising in resistance from EIS characterization, indicating a good stability both on active sites and support. The MEAs using such IrOx/Ti as the anode were prepared by conventional CCM method, and the Pt loading (20% wt. Pt/C) in cathode was reduced to about 0.2mg/cm2. The anode with various Ir loading and membrane with different thickness were determined at atmosphere under 80 degrees. The electrolyser with Ir loading 0.13mg/cm2 in anode, Pt loading 0.18mg/cm2 in cathode and N212 membrane achieved 1.623V at 1A/cm2 and 1.778V at 2A/cm2, which exhibits a good performance with such low noble metal loading. Further decreasing the Ir loading in anode catalyst layer will lead to performance degradation, but the rising of voltage was only 50-60mv at 1A/cm2 when Ir was reduced to 0.08 mg/cm2. With respect of the durability of MEAs, an electrolyser using Nafion 117 as the membrane, and 0.08 mg /cm2 Ir in MEA was tested under 1A/cm2. After 850h, the voltage increased from 1.79V to 1.83V, and the degradation rate is below 50μV/h, this suggests a good stability of supported Iridium oxohydroxides, and the stability of electrolyser with N212 membrane is now ongoing. Figure 1

One-dimensional structure with high surface area, oriented charge transfer and facile mass transfer is a promising solution to reduce the loading amount of noble metal as catalyst in PEM water electrolysis. In this work, TiO2 nanotube array (TNTA) is applied to decorate IrO2 to construct highly ordered 1-D electrode. Firstly, TNTA, synthesized by anodizing in hydrofluoric solution, is used as substrate as its great stability in both acidic and oxidizing environment. By considering the insufficient conductivity of TiO2, modification, including calcining in hydrogen or cathodic polarization, is conducted to improve its conductivity. Both the method can reduce part of Ti4+ to Ti3+ to achieve improving conductivity. By comparison, cathodic polarization treated TNTA performs higher carrier density without hydrogen brittleness. Iridium oxide is electro-deposited onto TNTA. By increasing the conductivity of TNTA, IrO2/reduced-TNTA shows a three order of magnitude higher OER activity than IrO2/pristine-TNTA. Reduced-TNTA is predictably unstable in oxidizing environment. By increasing the coverage of IrO2, the stability of IrO2/reduced-TNTA increase as the surface layer of IrO2 will isolate the unstable reduced-TNTA from oxidizing species generated during OER to prevent decrease of substrate conductivity. Secondly, as a compact layer of IrO2 on the surface of reduced-TNTA is essential to its stability, TNTA is used as sacrificed template to fabricate one-dimensional IrO2 nano-array electrode. By removing TNTA, the mass-ratio surface area of IrO2 is double. Hot-press treatment followed by two-step etching of TNTA is conducted to transfer the IrO2 nano-array to Nafion membrane. With 1500 nanometer long TNTA prepared in hydrofluoric sodium solution is used as template, iridium oxide nanotube arrays and hollow nanowire arrays is achieved by applying different scan rate in deposition. The nanotube array is consisted with intact nanotube and porous nanotube surrounding it, which is corresponding to the structure of template. A mechanism of the scan rate controlling deposition is proposed. The deposition of iridium oxide on TNTA can be divided into the deposition near the mouth and the deposition along the tube wall. With continuously depositing, the concentration in the nanotube will gradually decrease and form a gradient. While the concentration near the mouth keep constant as its sufficient diffusion of precursor from the bulk electrolyte. So, if the duration of deposition in a single cycle is long enough, iridium will be preferentially deposited at the mouth of TNTA. And once the cover layer on the top surface is formed, the diffusion into the tube will be limited and further deposition will only lead to the thickening of the cover layer. As a result, with slow scan rate, the cover layer will shell the mouth of TNTA before the constant layer of iridium is deposited throughout the template, and results in nanotube arrays. And with fast scan rate, the deposition is more even, and the hollow nanowire arrays is obtained. Electrochemical study shows that the current density of iridium oxide nano-arrays in OER is positive correlated to its surface area, and it shows better performance of mass-ratio activity than commercial iridium oxide nanoparticle as anode in MEA.

The transmission characteristics of a 2-dimensional (2D) square lattice photonic crystal waveguide (PCW) with an additional perpendicular line defect are investigated by the plane wave expansion method (PWM) and the finite difference time domain (FDTD) method. From simulation results and theoretical analysis, it is found that changing the refractive index of the dielectric columns in the additional line defect can effectively adjust the light transmission through the PCW. The working mechanism can be boiled down to either the excitation of a single defect mode or the contra-directional coupling of the two defect modes. And this mechanism can be used to design other tunable optical devices.

The impact factors were explored to determine the horizontal positional relationship between the umbilicus and the 2nd lumbar spinal process in adults and to verify the accuracy of the localization of Shenshu (BL 23) via the umbilicus. The position of the umbilicus and the 2nd lumbar spinal process was measured in 100 participants and the data were analyzed through SPSS 20.0 software. It was found that the umbilicus and the 2nd lumbar process were not positioned horizontally. The positional relationship of these two sites was not apparently correlated with gender, age, body weight, body height, BMI, waistline and discomfort of lumbar region. The umbilicus was commonly and posteriorly projected on the site between the 4th and 5th lumbar vertebra. It is explained that the localization of Shenshu (BL23) via the umbilicus is not accurate.

With Jianyu (LI 15), Jianliao (TE 14) and Jianzhen (SI 9) for instance, the three acupoints used to treat shoulder bi syndrome by needle knife, and through traceability and researching constant structure, the acupoint position, insertion trace, manipulation and clinical significance of needle knife medicine were discussed. Accurate position is one of the characteristics of acupoints selection of needle knife medicine. As for the acupoints selection method, the mean of body surface localization is always used. The phanerous or palpable bone processes, muscles and tendons are taken as positioning marks; pressing areas where appear sour, numb, or distensible and other sensations is considered as the principle of press positioning. So acupoints position method is the combination of observation and palpation. Different insertion methods can effectively relieve the accretive bursae synovialis, tendon, joint capsule and the compressed nerve, so that shoulder bi syndrome is relieved.

Since enterprises installed the computer-based accounting information system, the environment for financial accounting and accounting control therein have changed dramatically, and the accounting data processing has been accelerated. In addition, the financial accounting has witnessed great improvement in its accuracy and reliability. Meanwhile, the huge differences between the artificial accounting information processing and computer-based accounting information processing have brought new problems and challenges to the internal control. As a result, it is of great significance to study and enhance the internal control of accounting information processing system under the computer environment.

We study the imaging process for a photonic crystal slab lens with a surface defect by the finite-difference-time-domain (FDTD) method. We demonstrate an odd surface mode with a specific dispersion curve in this system. The dispersion curve has an extreme point, which is corresponding to a slow light. If the working frequency is chosen at this extreme point the subwavelength resolution of image will be enhanced. Moreover, the subwavelength resolution of image is very sensitive to the position of this extreme point in the dispersion diagram. Longer interaction time and better field distribution may give a qualitative physical understanding for the enhancement of imaging quality.

Hydrogen production from catalytic steam reforming of model compound acetic acid of bio-oil was investigated under atmospheric pressure in a bench-scale fixed-bed reactor and a series of Ni catalysts supported on ceria–zirconia were prepared via co-precipitation with varying contents of Ni and Ce loading. Simultaneously, effects of reaction temperature, and molar ratio of steam to carbon fed, liquid hourly space velocity (LHSV) were also explored. The results indicated that the 12Ni/7.5Ce-Zr-O catalyst was superior to other catalysts with different Ni and Ce loading. At T = 650 °C, S / C = 3, LHSV = 2.8 h −1 , the reaction over Ni catalyst supported on ceria–zirconia turns out that the highest activity and the maximum hydrogen selectivity of 83.4% and the minimum methane selectivity of 0.39% are attained. Besides, Ni catalyst supported on ceria–zirconia shows good stability during 25 h continuous reaction without any deactivation.

A tunable power splitter is proposed in a two-dimensional photonic crystal with a line defect composed of two rows rods, whose refractive index can be controlled artificially. This device is working at a self-collimation frequency. The finite difference time domain (FDTD) simulations show that the beam splitting ratio is a function of the refractive index of line defect. If the index varies from 3.0 to 3.5, the beam intensity of one output port decreases from 87.2% (normalized by the input beam) to 0 while the intensity of the other output port increases from 8.98% to 100% monotonically. The coupling between the self-collimation mode and the defect mode is used to explain the working mechanism.

The aim of the present study was to make use of the artificially induced aging model cardiomyocytes to further investigate potential anti-aging-associated cellular diastolic dysfunction effects of EGB761 and explore underlying molecular mechanisms. Cultured rat primary cardiomyocytes were treated with either D-galactose or D-galactose combined with EGB761 for 48 h. After treatment, the percentage of cells positive for SA-β-gal, AGEs production, cardiac sarcoplasmic reticulum calcium pump (SERCA) activity, the myocardial sarcoplasmic reticulum calcium uptake, and relative protein levels were measured. Our results demonstrated that in vitro stimulation with D-galactose induced AGEs production. The addition of EGB761 significantly decreased the number of cells positive for SA-β-gal. Furthermore, decreased diastolic [Ca2+]i, curtailment of the time from the maximum concentration of Ca2+to the baseline level and increased reuptake of Ca2+stores in the SR were also observed. In addition, the level of p-Ser16-PLN protein as well as SERCA was markedly increased. The study indicated that EGb761 alleviates formation of AGEs products on SERCA2a in order to mitigate myocardial stiffness on one hand; on other hand, improve SERCA2a function through increase the amount of Ser16 sites PLN phosphorylation, which two hands finally led to ameliorate diastolic dysfunction of aging cardiomyocytes.

We show a dual-beam-reflection phenomenon for a Gaussian beam illuminating a hybrid structure of a dielectric waveguide and photonic crystal (WG-PC) inside the photonic band gap by numerical simulations. One reflection beam has a giant negative lateral shift, but the other has a positive lateral shift. The finite-difference time-domain (FDTD) simulations show that this phenomenon has a time delay effect and comes from the leaky surface mode of the hybrid structure. Field profile of the leaky mode demonstrates a strong localized stationary field in the higher dielectric medium. Furthermore, the maximum lateral shift is almost two times of the waist of the incident beam.

Considering the non-uniformity of the flow velocity distribution in fluid-conveying pipes caused by the viscosity of real fluids, the centrifugal force term in the equation of motion of the pipe is modified for laminar and turbulent flow profiles. The flow-profile-modification factors are found to be 1.333, 1.015–1.040 and 1.035–1.055 for laminar flow in circular pipes, turbulent flow in smooth-wall circular pipes and turbulent flow in rough-wall circular pipes, respectively. The critical flow velocities for divergence in the above-mentioned three cases are found to be 13.4%, 0.74–1.9% and 1.7–2.6%, respectively, lower than that with plug flow, while those for flutter are even lower, which could reach 36% for the laminar flow profile. By introducing two new concepts of equivalent flow velocity and equivalent mass, fluid-conveying pipe problems with different flow profiles can be solved with the equation of motion for plug flow.

The effective, stable and low cost anodic catalyst or electrode for OER is still a big challenge for water electrolysis in acid condition. Many efforts were devoted to develop suitable supports for depositing Ir and Ru metal or oxides to obtain an enhanced activity or stability for OER. In this work, self-doped TiO2 nanotube arrays (sTNTA) which has a higher electrical conductivity than TNTA was chosen as the support, it was fabricated by a simple electrochemical reduction of TNTA in neutral electrolyte. Then IrO2 nano particles were deposited on it by electrochemical pulse deposition method to form an OER electrode. The cyclic voltammetric behavior, electrical conductivity and micro structure of sTNTA prepared under different reduction potential were characterized to obtain an optimal performance for depositing IrO2, the OER activity and stability of new electrode were also determined. The obtained sTNTA shows an increased current in CV test than TNTA, which is due to the increased carrier density. As the reduction potential move to more negative level (from -1.5V to -1.9V vs. Ag/AgCl), the carrier density was found increased for 3 order of magnitudes, and XPS characterization confirm the existence of Ti3+ which formed in electrochemical reduction process, and sTNTA prepared under -1.9V shows the best electrochemical performance. After depositing IrO2, IrO2/sTNTA exhibited higher OER activity than IrO2/TNTA, according to the CV and EIS test, this can be ascribed to the enhanced electrical conductivity of support, which results in more active sites and lower charge transfer resistance of the electrode. The OER stability of IrO2/sTNTA was tested under constant current of 5mA/cm2, IrO2/TNTA and IrO2/Ti which prepared by electrochemical deposition on Ti foil were also tested for comparison. A higher stability of IrO2/sTNTA electrode was observed than the other two electrodes, and this may be due to the interaction between the IrO2 and sTNTA. XPS studies show a clear peak shift of Ir4f in IrO2/sTNTA, indicating a possible lower valence of Iridium, and such lower oxidation state of Iridium may be response for the enhanced OER stability. In summary, our studies have demonstrated the possibility of self-doped TNTA as a potential anodic support for OER. Comparing with the TNTA, the high electrical conductivity of sTNTA facilitate the charge transfer rate and thus improve the OER activity. Additionally, a higher OER stability was observed for IrO2/sTNTA electrode, and this may be due to possible catalyst-support interactions between IrO2 and self-doped TNTA. Figure 1

Background: Pancreatic cancer is characterized by high resistance to chemotherapy. Such chemoresistance can be mediated by multidrug resistance proteins (MRPs), breast cancer resistance protein (BCRP), and MDR1 P-glycoprotein. However, the contribution of individual MRP isoforms to chemoresistance in pancreatic carcinoma is unclear. We studied ATP-binding cassette (ABC) transporter expression in human pancreatic carcinoma cell lines as compared to primary pancreatic duct cells, and analyzed the MRP expression profile in 5-fluorouracil-resistant cells. Methods: Transporter expression was analyzed by quantitative and qualitative RT-PCR, by immunoblot, and chemoresistance by cytotoxicity assay. Results: Primary pancreatic duct cells expressed MRP1, MRP3, MRP4, and MRP5, but not MRP2 mRNA. The established carcinoma cell lines expressed MRP1, MRP4, and MRP5, most of them also MRP2, MRP3, MRP7, and BCRP, but none contained detectable amounts of MRP6, MRP8, or MRP9 mRNA. Immunoblot analyses demonstrated presence of MRP1, MRP4, and MRP5 protein in all, but MRP3 and BCRP protein only in some of these cells. Compared to parental Capan-1 cells, Capan-1 cells with acquired chemoresistance towards 5-fluorouracil showed an upregulated mRNA and protein expression of MRP3, MRP4, and MRP5. In addition, silencing of MRP5 by RNA interference resulted in enhanced sensitivity of parental Capan-1 cells towards 5-fluorouracil cytotoxicity. Conclusion: MRP3, MRP4, and MRP5 are upregulated in 5-fluorouracil-resistant cells, and MRP5 contributes to 5-FU resistance in pancreatic carcinoma cells.

Boron solubility in the as-cast and solution treated martensite of Fe-Cr-B cast irons, containing approximately 1.35 wt.% of boron, 12 wt.% of chromium, as well as other alloying elements, has been investigated using conventional microanalysis. The significant microstructural variations after tempering at 750 degreesC for 0.5-4 h, compared with the original as-cast and solution treated microstructures, indicated that the matrix consisted of boron and carbon supersaturated solid solutions. The boron solubility detected by electron microprobe was between 0.185-0.515 wt.% for the as-cast martensite and 0.015-0.0589 wt.% for the solution treated martensite, much higher than the accepted value of 0.005 wt.% in pure iron. These remarkable increases are thought to be associated with some metallic alloying element addition, such as chromium, vanadium and molybdenum, which have atomic diameters larger than iron, and expand the iron lattice to sufficiently allow boron atoms to occupy the interstitial sites in iron lattice. (C) 2002 Elsevier Science B.V. All rights reserved.

Thermoelectric generators(TEG) is considered as a promising technology for utilizing low-grade thermal energy, while the cooling system plays a key role in TEG. A two-dimensiona l numerical model of jet impinging jet cooling is developed based on the commercial code FLUENT. The influences of several factors on thermoelectri c conversion efficiency are observed. The factors include the air inlet velocity, the TE legs height, and the fins. The results show that the air inlet velocity is one of the most important factors for the cooling performance, the convection heat transfer efficiency and conversion efficiency significantly increases with the improvement of the air velocity. Higher TE legs height can influence the TEG performance effectively, 5~7mm legs height is recommended in this study. And fins can effectively increase the TEG efficiency, the thermoelectric conversion efficiency with 1mm×1mm fins is 6.73% which was 2.4 times higher than without fins.

A series of new two semiconductor catalysts, Co3O4/CeO2, were prepared by glycine-nitrate combustion method for photocatalytic reduction of carbon dioxide to produce methanol and ethanol under visible light (λ>400 nm) irradiation. The catalysts were characterized by BET, UV-vis spectra, XRD, SEM, PL, and XPS and the results indicated that the catalyst with 5 wt.% of Co3O4has the highest yield among all kinds of tests with the methanol yield of 1.52 μmol·g−1·h−1and the ethanol yield of 4.75 μmol·g−1·h−1, which are about 2.34 and 1.71 times as large as those of CeO2. However, methanol and ethanol can hardly be detected for Co3O4under the same condition because of its too narrow band gap. The improvement of the photoreduction activity of Co3O4doped CeO2was caused by the separation of electron-hole pairs of Co3O4/CeO2and charge transfer between Co3O4and CeO2, mimicking the Z-scheme in photosynthesis.

A face recognition system can be deceived by photos, mimic masks, mannequins and etc. And with the advances in the 3D printing technology, a more robust face liveness detection method is needed. In this paper, a gradient-based multispectral method has been proposed for face liveness detection. Based on two spectral bands, the developed method is tested for the classification of genuine faces and common disguised faces. A true positive rate of 96.7% and a true negative rate of 97% have been achieved. The performance of the method is also tested when face rotation occurs. The contributions of this paper are: First, a gradient-based multispectral method has been proposed. Except for the reflectance of the skin regions, the reflectance of other distinctive regions in a face are also considered in the developed method. Second, the method is tested based on a dataset with both planar photos and 3D mannequins and masks. The performance on different face orientations is also discussed.

The spatial distribution of the eutectic M B borides in Fe-Cr-B cast irons was predicted using strain models. The spatial distribution of the eutectic M B borides was revealed by scanning electron microscopy (SEM) using deep-etched specimens. It was shown that the models are helpful for the prediction of mechanical properties, as well as interpretation of mechanical property difference of Fe-Cr-B cast irons. The results show that the models are successful in the prediction of the spatial distribution of the eutectic M B borides.

Density functional theory (DFT) is used to study the NO reduction by H 2 on Au 4 + and Au 4 clusters. The reaction mechanism is explored along two possible entrances: one involves the complexes of the clusters with H 2 and the other is related to the complexes of the clusters with NO. In all cases, it is found that the catalytic cycle involves two sequential elementary steps: the rupture of the H-H bond in H 2 and the formation of H 2 O and N 2 O molecule. The calculated results show that the reaction mediated by Au 4 + is energetically most favorable compared to that promoted by Au4, indicating that the charge state of Au clusters plays an essential role for the catalyzed NO reduction. The present theoretical study rationalizes the early experimental findings well and enriches our understanding of the catalytic NO reduction by Au-based catalysts.

Pancreatic cancer is characterized by high resistance to chemotherapy. Such chemoresistance can be mediated by multidrug resistance proteins (MRPs), breast cancer resistance protein (BCRP), and MDR1 P-glycoprotein. However, the contribution of individual MRP isoforms to chemoresistance in pancreatic carcinoma is unclear. We studied ATP-binding cassette (ABC) transporter expression in human pancreatic carcinoma cell lines as compared to primary pancreatic duct cells, and analyzed the MRP expression profile in 5-fluorouracil-resistant cells.Transporter expression was analyzed by quantitative and qualitative RT-PCR, by immunoblot, and chemoresistance by cytotoxicity assay.Primary pancreatic duct cells expressed MRP1, MRP3, MRP4, and MRP5, but not MRP2 mRNA. The established carcinoma cell lines expressed MRP1, MRP4, and MRP5, most of them also MRP2, MRP3, MRP7, and BCRP, but none contained detectable amounts of MRP6, MRP8, or MRP9 mRNA. Immunoblot analyses demonstrated presence of MRP1, MRP4, and MRP5 protein in all, but MRP3 and BCRP protein only in some of these cells. Compared to parental Capan-1 cells, Capan-1 cells with acquired chemoresistance towards 5-fluorouracil showed an upregulated mRNA and protein expression of MRP3, MRP4, and MRP5. In addition, silencing of MRP5 by RNA interference resulted in enhanced sensitivity of parental Capan-1 cells towards 5-fluorouracil cytotoxicity.MRP3, MRP4, and MRP5 are upregulated in 5-fluorouracil-resistant cells, and MRP5 contributes to 5-FU resistance in pancreatic carcinoma cells.

Face liveness detection aims to distinguish genuine faces from disguised faces. Most previous works under visible light focus on classification of genuine faces and planar photos or videos. To handle the three-dimensional (3-D) disguised faces, liveness detection based on multispectral images has been shown to be an effective choice. In this paper, a gradient-based multispectral method has been proposed for face liveness detection. Three feature vectors are developed to reduce the influence of varying illuminations. The reflectance-based feature achieves the best performance, which has a true positive rate of 98.3% and a true negative rate of 98.7%. The developed methods are also tested on individual bands to provide a clue for band selection in the imaging system. Preliminary results on different face orientations are also shown. The contributions of this paper are threefold. First, a gradient-based multispectral method has been proposed for liveness detection, which considers the reflectance properties of all the distinctive regions in a face. Second, three illumination-robust features are studied based on a dataset with two-dimensional planar photos, 3-D mannequins, and masks. Finally, the performance of the method on different spectral bands and face orientations is also shown in the evaluations.