Search

Scene text detection methods have achieved significant progresses. However, stack-omnidirectional text dilemma, under-segmentation of very close text words, and over-segmentation of arbitrary-shape long text lines, are still main challenges. Motivated by these problems, we proposed a two stage method called omnidirectional pyramid mask proposal text detector (OPMP). OPMP removes anchor mechanism that requires heuristic non-maximum suppress processing. Instead, it uses an effective pyramid lengthwise and sidewise residual sequence modeling method to produce arbitrary-shape proposals. To accurately extract the features of text shape, OPMP enhances the backbone layers by a multiple arbitrary-shape fitting mechanism. Finally, a multi-grain text classification module is proposed, which reclassifies each text region robustly. Comprehensive ablation studies demonstrate the effectiveness of each proposed component. In addition, experiments on various benchmarks, including ICDAR2015, MLT, MSRA-TD500, CTW1500, and Total-text, show that our method outperforms previous state-of-the-art methods.

This paper proposes a general analysis model and an improved arm control of modular multilevel converter (MMC) under asymmetric operation conditions. This general analysis model is established as the arm voltage equations with equivalent impedance parameters, and is applicable for characteristics analysis under both submodule (SM) fault condition and asymmetric arm parameter condition. Based on the analysis model, the equivalent circuits of the circulating current and the ac side current are obtained, and then the asymmetric characteristics are performed. According to the analysis results, an improved arm control of MMC is proposed, which is aimed at balancing the dc components in upper and lower arm currents, as well as suppressing the fundamental and double frequency circulating currents. Besides, the dynamic response of the proposed arm control is significantly improved by embedding an arm voltage inertial feedback control. Simulations on PSCAD/EMTDC verify the correctness of the theoretical analysis and the validity of the proposed control strategy.

Objectives To investigate whether dosiomics can benefit to IMRT treated patient’s locoregional recurrences (LR) prediction through a comparative study on prediction performance inspection between radiomics methods and that integrating dosiomics in head and neck cancer cases. Materials and Methods A cohort of 237 patients with head and neck cancer from four different institutions was obtained from The Cancer Imaging Archive and utilized to train and validate the radiomics-only prognostic model and integrate the dosiomics prognostic model. For radiomics, the radiomics features were initially extracted from images, including CTs and PETs, and selected on the basis of their concordance index (CI) values, then condensed via principle component analysis. Lastly, multivariate Cox proportional hazards regression models were constructed with class-imbalance adjustment as the LR prediction models by inputting those condensed features. For dosiomics integration model establishment, the initial features were similar, but with additional 3-dimensional dose distribution from radiation treatment plans. The CI and the Kaplan–Meier curves with log-rank analysis were used to assess and compare these models. Results Observed from the independent validation dataset, the CI of the model for dosiomics integration (0.66) was significantly different from that for radiomics (0.59) (Wilcoxon test, p = 5.9 × 10 - 31 ). The integrated model successfully classified the patients into high- and low-risk groups (log-rank test, p = 2.5 × 10 - 02 ), whereas the radiomics model was not able to provide such classification (log-rank test, p = 0.37 ). Conclusion Dosiomics can benefit in predicting the LR in IMRT-treated patients and should not be neglected for related investigations.

Although melatonin has been extensively studied in animal reproduction, the mechanism of melatonin in puberty remains elusive. This study was designed to explore the effect of intraperitoneal administration of melatonin on puberty onset in female mice. The injection of melatonin into postnatal days 10 mice at a dose of 15 mg/kg accelerated the puberty onset in mice. Mechanistically, there was no difference in physical growth and serum Leptin levels after melatonin administration. Meanwhile, the serum levels of reproductive hormones involved in hypothalamic-pituitary-ovarian axis, such as FSH and estrogen level in serum were increased. The mRNA levels of GnRH and GnRHr were not affected by melatonin, while the expressions of FSHβ in pituitary and Cyp19a1 in ovary were significantly up-regulated. In addition, melatonin still promoted FSH synthesis after ovariectomy. Furthermore, the enhanced activity of ERK1/2 signaling verified that the expression of FSHβ increased in pituitary. We confirmed that melatonin promoted the FSH synthesis in pituitary, thereby increased serum estrogen levels and ultimately accelerated puberty onset. However, these effects of melatonin may be pharmacological due to the high dose. This study would help us to understand the functions of melatonin in pubertal regulation comprehensively.

Underwater range-gated laser imaging (URGLI) still has some problems like un-uniform light, low brightness and contrast. To solve the problems, a variant of adaptive histogram equalization called contrast limited adaptive histogram equalization (CLAHE) is proposed in this paper. In experiment, using the CLAHE and HE to enhance the images, and evaluate the quality of enhanced images by peak signal to noise ratio (PSNR) and contrast. The result shows that the HE gets the images over-enhanced, while the CLAHE has a good enhancement with compressing the over-enhancement and the influence of un-uniform light. The experimental results demonstrate that the CLAHE has a good result of image enhancement for target detection by underwater range-gated laser imaging system.

Multiplanar tubular KK-joints are one of the most popular joint types in offshore structures, which are exposed to cyclic loads and fatigue damages. In this paper, a fatigue prediction method based on scaled model test is proposed. First, a scaled KK-joint, including deviations for model simplification, was designed based on sensitive analysis and similarity analysis. Then, static and fatigue tests on the scaled model under axial loading were performed, by which hot spot stress (HSS) distributions and the maximum HSS were recorded. From the test, the fatigue crack initiates from the location of the maximum HSS and propagates along the weld toe. Finally, the maximum HSS of original KK-joint was deduced by finite element analysis (FEA), and then, the fatigue life was predicted accordingly and compared with the rule-based result.

The bending characteristics of distributed feed back fiber lasers were investigated.Firstly,a theoretical model of fiber lasers output wavelength and bending radius was presented.It shows a linear correlation between the output wavelength variation of fiber lasers and the inverse square of the bending radius,and as the bending radius increases,the Bragg wavelength experiences a blue shift.In corresponding experiments,3 fiber lasers were used to test output wavelength with bending radii between 10 mm and 90 mm;these experiments demonstrate the accuracy of the theoretical analysis.Secondly,fiber lasers output power and bending radius were studied.One fiber laser was used as reference to eliminate environmental interference.The experiment shows fluctuation ranges of fiber lasers output power are less than 2 dB during the experimental time in 32 days.Finally,the slope efficiency is investigated,showing excellent bending insensitivity.The results provide an important reference for DFB fiber lasers in applications of bending sensing.

An electro-optic variable optical attenuator in silicon-on-insulator is designed and fabricated. A series Structure is used to improve the device efficiency Compared to the attenuator in the single p-i-n diode Structure in the same modulating length, the attenuation range of the device in the series structure improves 2-3 times in the same injecting current density, while the insertion loss is not affected. The maximum dynamic attenuation of the device is greater than 30 dB. The response frequency is obtained to be about 2 MHz.

为更适用于批量生产激光投影仪，需要一种具有结构紧凑、性能稳定与低成本等特点的绿色激光器作为绿光光源。采用光胶合的Nd：YVO4/PPMgOLN晶体棒作为增益介质和倍频晶体，结合一种新型焊接封装工艺，最终得到了一种紧凑型微片阵列激光器。该激光器3个激光光束共得到了223.7 mW的功率输出，每个激光光束都有良好的光斑分布和稳定性，且激光器包含两个内置TEC在内具有47 mm35 mm25 mm的紧凑尺寸。激光器连续工作2.5 h波动小于2.5%。各项性能可以充分满足激光投影仪绿光光源的要求，并便于低成本批量生产与后期维护。

Cytochrome P450 2A6 (CYP2A6) is an enzyme involved in the metabolism of some tobacco carcinogens, which is an important risk factor of lung cancer. Among CYP2A6 allelic variants, CYP2A6*4 presents a whole gene deletion that accounts for the majority of poor metabolizer. In this study, a meta-analysis was performed to assess the association between CYP2A6*4 and risk of lung cancer. Literature searches were conducted to identify peer-reviewed manuscripts published up to December 20, 2012. Pooled odds ratios (ORs) and 95 % confidence intervals (95 % CIs) were calculated in a fixed-effects model and a random-effects model when appropriate. Eight eligible studies with 3,203 lung cancer cases and 2,839 controls were included in this study. Overall, no significant association was observed in CYP2A6*4 with the risk of lung cancer under any genetic model for all samples after correction. However, subgroup analysis showed that significant associations were observed in Asian with pooled OR (95 %CI) of 0.761 (0.672–0.861) for allele comparison, 0.769 (0.668–0.886) for dominant model, and 0.522 (0.359–0.760) for recessive model. Furthermore, after stratifying Asian samples according to smoking status, significant associations were only observed in smokers with pooled OR (95 %CI) of 0.713 (0.607–0.838) for allele comparison, 0.720 (0.596–0.869) for dominant model, and 0.444 (0.275–0.715) for recessive model. This meta-analysis suggests that the CYP2A6*4 polymorphism was associated with susceptibility of lung cancer for smokers in Asian. The whole gene deletion of CYP2A6 might decrease the risk of tobacco-related lung cancer in Asian.

Coastal surveillance is very important because it is useful for search and rescue, illegal immigration, or harbor security and so on. Furthermore, range estimation is critical for precisely detecting the target. Range-gated laser imaging sensor is suitable for high accuracy range especially in night and no moonlight. Generally, before detecting the target, it is necessary to change delay time till the target is captured. There are two operating mode for range-gated imaging sensor, one is passive imaging mode, and the other is gate viewing mode. Firstly, the sensor is passive mode, only capturing scenes by ICCD, once the object appears in the range of monitoring area, we can obtain the course range of the target according to the imaging geometry/projecting transform. Then, the sensor is gate viewing mode, applying micro second laser pulses and sensor gate width, we can get the range of targets by at least two continuous images with trapezoid-shaped range intensity profile. This technique enables super-resolution depth mapping with a reduction of imaging data processing. Based on the first step, we can calculate the rough value and quickly fix delay time which the target is detected. This technique has overcome the depth resolution limitation for 3D active imaging and enables super-resolution depth mapping with a reduction of imaging data processing. By the two steps, we can quickly obtain the distance between the object and sensor.

In this research, the artificial intelligent method based human motion pattern recognition for surface electromyographic (EMG) signal is proposed. As the EMG signal is a measurement of anatomical and physiological characteristic of the given muscle, the macroscopical movement patterns of the human body can be classified and recognized. By using the technology of wavelet packet transformation, the high-frequency noises can be eliminated effectively and the characteristics of EMG signals can be extracted. Auto-regressive model is adopted to effectively simulate the stochastic and non-stationary time sequences using a series of auto-regressive coefficients with a typical order. Artificial neural network is utilized to distinguish the different force levels in the game of arm wrestling. The efficiency of the proposed methods are proved by experiment results.

Human-Computer Interface (HCI) is very important to improve the efficiency of robot control to support the complex tasks. It is valuable to develop recognition techniques by which compute can be capable of handling human’s continuous gesture effectively. The present work seeks to model human motions in a manner for leaning and recognition. Toward this end,a novel HCI approach based on hidden Markov model(HMM) is proposed to model semantically meaningful human movements by a sequence of camera images. As a maximum likelihood recognizer, HMM is employed to learn each motion class. Finally, experimental results show the performance of the proposed approach with 1000 pictures.

In order to get high efficiency of code design and convenience of system maintenance, this paper applies the most advanced method, unified modeling language (UML) and model-driven architecture (MDA), on the development of the embedded system based on the MSC1212. Firstly, using Matlab/stateflow toolbox to model and simulate the complex event driven system in the embedded system introduced in this paper. Then using stateflow coder translated the state chart into high efficiency C code. By making some simple replacement to the blank C function with the low-layer function of the target system based MSC1212, the code rapid generation of the embedded software is finished. It has been proved by experiment that, the automatically code generation introduced in this paper works well in the embedded target system. It has good performance and high code efficiency than those code generated by hand.

A novel ambient intelligence interactive technique are introduced here. First, the mechanism of our AmI system and several idiographic technique issue is expounded stepwise. For providing personalized service an algorithm of face recognition based on HMM is researched and the experimental results show efficient. In this way the user can be identified by the method of face recognition, and the system with a LCS can provide personalized services. Otherwise, we adopt the wavelet analysis and template matching technique to estimate the user's basic face orientation, thus the video projector is driven to rotate toward the same direction. In sum, the study described here is likely to be used in intelligent meeting or medical treatment environment.

The near infrared spectroscopy for the noninvasive determination of blood compositions has attracted significant interest in recent years. Lately the Dynamic Spectroscopy (DS), a new device for artery blood compositions detecting is introduced. It can eliminate the interferences of individual discrepancy and measurement conditions. But acquisition of the spectrum meeting the accuracy of blood compositions analysis is a bottleneck to be resolved. In this article, a new acquisition device of dynamic spectrum is present. On the hardware side, logarithmic amplifier is used to convert the photoelectric pulse wave to the logarithmic pulse wave. This modification compresses the dynamic range of the system and reduced the cost of the signal picking device. On the software side, based on the fast Fourier Transform (FFT) algorithm, the calculating of the peak to peak value of the photoelectric pulse wave in the time domain is converted to the extracting of the fundamental components of the logarithmic pulse wave. Due to the plus function of FFT to the signal to noise ratio (SNR), the resolution of the spectrum analysis is improved.

Part perfusion hyperthermia is an effective approach to tumor curing through which the tumor cells are killed by the high temperature, which is created by medicament perfuse to the pathological change area. Since the human body is a complicated system of heat exchange. It is difficult to obtain its accurate mathematic model and accurately measure the temperature of cancer tissue applying the conventional method. The paper proposed a novel intelligent modeling and measuring scheme based on RBF neural network, which can conduct the accurate measurement of the temperature of cancer tissue. The neural network model is designed and the training algorithm with high-precision of learning is formulated based on neural networks techniques. Compared with the conventional scheme, the proposed intelligent scheme has the following advantage: good robustness and adaptive ability for different cancer tissues, simple implementation and high-precision. A lot of numerical simulations have been conducted, and simulation results have shown that the method is feasible and efficient.

Based on the large cross-section single-mode rib waveguide condition, the total internal reflection and the free-carrier plasma dispersion effect, a silicon-on-insulator (SOI) 2 × 2 symmetric optical waveguide switch with a transverse injection structure has been proposed and fabricated, in which the SOI technique utilizes silicon and silicon dioxide thermal bonding and back-polishing. The device performance is measured at the wavelength of 1.3 μm. It shows that the crosstalk and insertion loss are less than - 18.1 and 4.8 dB, respectively, at an injection current of 60 mA, and response time is 110 ns.

Semiconductor optical switches utilizing carrier-induced refractive index change have tremendous potential for the application of optical processing because of their small size , single-mode operation , polarizationindependence and wavelength-insensitivity. These switches are generally based on total internal reflection , and have a requirement of precise control of reflection interface1, 2. The requirement causes a difficulty in developing silicon switches of the same kind. Because silicon is an indirect-gap semiconductor , the injected carriers in it diffuse a long distance due to their long lifetime. They can not be restricted within a specific region in the normal structure like what has been taken in InGaAsP or A1GaAs. Although some authors have reported that attempt to make use of the free-carrier effect in silicon for optical switching, the developed switches are all based on mode-interference other than internal reflection3, 4. These switches are devoid of the advantages mentioned above. From the viewpoint of practical application , a kind of silicon switch with these advantages is more essential.