Retinal vascular morphological characteristics in diabetic retinopathy: an artificial intelligence study using a transfer learning system to analyze ultra-wide field images

AIM:To investigate the morphological characteristics of retinal vessels in patients with different severity of diabetic retinopathy(DR)and in patients with or without diabetic macular edema(DME).METHODS:The 239 eyes of DR patients and 100 eyes of healthy individuals were recruited for the study.The severity of DR patients was graded as mild,moderate and severe non-proliferative diabetic retinopathy(NPDR)according to the international clinical diabetic retinopathy(ICDR)disease severity scale classification,and retinal vascular morphology was quantitatively analyzed in ultra-wide field images using RU-net and transfer learning methods.The presence of DME was determined by optical coherence tomography(OCT),and differences in vascular morphological characteristics were compared between patients with and without DME.RESULTS:Retinal vessel segmentation using RU-net and transfer learning system had an accuracy of 99%and a Dice metric of 0.76.Compared with the healthy group,the DR group had smaller vessel angles(33.68±3.01 vs 37.78±1.60),smaller fractal dimension(Df)values(1.33±0.05 vs 1.41±0.03),less vessel density(1.12±0.44 vs 2.09±0.36)and fewer vascular branches(206.1±88.8 vs 396.5±91.3),all P<0.001.As the severity of DR increased,Df values decreased,P=0.031.No significant difference between the DME and non-DME groups were observed in vascular morphological characteristics.CONCLUSION:In this study,an artificial intelligence retinal vessel segmentation system is used with 99%accuracy,thus providing with relatively satisfactory performance in the evaluation of quantitative vascular morphology.DR patients have a tendency of vascular occlusion and dropout.The presence of DME does not compromise the integral retinal vascular pattern.

High-resolution imaging of magnetic fields of banknote anti-counterfeiting strip using fiber diamond probe

Counterfeiting of modern banknotes poses a significant challenge,prompting the use of various preventive measures.One such measure is the magnetic anti-counterfeiting strip.However,due to its inherent weak magnetic properties,visualizing its magnetic distribution has been a longstanding challenge.In this work,we introduce an innovative method by using a fiber optic diamond probe,a highly sensitive quantum sensor designed specifically for detecting extremely weak magnetic fields.We employ this probe to achieve high-resolution imaging of the magnetic fields associated with the RMB 50denomination anti-counterfeiting strip.Additionally,we conduct computer simulations by using COMSOL Multiphysics software to deduce the potential geometric characteristics and material composition of the magnetic region within the anti-counterfeiting strip.The findings and method presented in this study hold broader significance,extending the RMB 50 denomination to various denominations of the Chinese currency and other items that employ magnetic anti-counterfeiting strips.These advances have the potential to significantly improve and promote security measures in order to prevent the banknotes from being counterfeited.

Real-time digital image stabilization based on regional field image gray projection

Digital image stabilization technique plays important roles in video surveillance and object acquisition.Many useful electronic image stabilization algorithms have been studied.A real-time algorithm is proposed based on field image gray projection which enables the regional odd and even field image to be projected into x and y directions and thus to get the regional gray projection curves in x and y directions,respectively.For the odd field image channel,motion parameters can be estimated via iterative minimum absolute difference based on two successive field image regional gray projection curves.Then motion compensations can be obtained after using the Kalman filter method.Finally,the odd field image is adjusted according to the compensations.In the mean time,motion compensation is applied to the even field image channel with the odd field image gray projection curves of the current frame.By minimizing absolute difference between odd and even field image gray projection curves of the current frame,the inter-field motion parameters can be estimated.Therefore,the even field image can be adjusted by combining the inter-field motion parameters and the odd field compensations.Finally,the stabilized image sequence can be obtained by synthesizing the adjusted odd and even field images.Experimental results show that the proposed algorithm can run in real-time and have a good stabilization performance.In addition,image blurring can be improved.

Evaluating the Potentials of PLSR and SVR Models for Soil Properties Prediction Using Field Imaging,Laboratory VNIR Spectroscopy and Their Combination

Pedo-spectroscopy has the potential to provide valuable information about soil physical,chemical,and biological properties.Nowadays,wemay predict soil properties usingVNIRfield imaging spectra(IS)such as Prisma satellite data or laboratory spectra(LS).The primary goal of this study is to investigate machine learning models namely Partial Least Squares Regression(PLSR)and Support Vector Regression(SVR)for the prediction of several soil properties,including clay,sand,silt,organic matter,nitrate NO3-,and calcium carbonate CaCO_(3),using five VNIR spectra dataset combinations(%IS,%LS)as follows:C1(0%IS,100%LS),C2(20%IS,80%LS),C3(50%IS,50%LS),C4(80%IS,20%LS)and C5(100%IS,0%LS).Soil samples were collected at bare soils and at the upper(0–30 cm)layer.The data set has been split into a training dataset 80%of the collected data(n=248)and a validation dataset 20%of the collected data(n=61).The proposed PLSR and SVR models were trained then tested for each dataset combination.According to our results,SVR outperforms PLSR for both:C1(0%IS,100%LS)and C5(100%IS,0%LS).For Soil Organic Matter(SOM)prediction,it achieves(R^(2)=0.79%,RMSE=1.42%)and(R^(2)=0.76%,RMSE=1.3%),respectively.The data fusion has improved the soil property prediction.The highest improvement was obtained for the SOM property(R^(2)=0.80%,RMSE=1.39)when using the SVR model and applying the second Combination C2(20% of IS and 80%LS).

Design of compact integrated diamond nitrogen-vacancy center quantum probe

An integrated quantum probe for magnetic field imaging is proposed,where the nitrogen–vacancy(NV)center fixed at the fiber tip is located on the periphery of flexible ring resonator.Using flexible polyimide(PI)as the substrate medium,we design a circular microstrip antenna,which can achieve a bandwidth of 140 MHz at Zeeman splitting frequency of 2.87 GHz,specifically suitable for NV center experiments.Subsequently,this antenna is seamlessly fixed at a three-dimensional-printed cylindrical support,allowing the optical fiber tip to extend out of a dedicated aperture.To mitigate errors originating from processing,precise tuning within a narrow range can be achieved by adjusting the conformal amplitude.Finally,we image the microwave magnetic field around the integrated probe with high resolution,and determine the suitable area for placing the fiber tip(SAP).

Infrared measurement of temperature field in coal gas desorption

In order to reveal the temperature change in coal gas desorption process,the temperature variation in coal gas desorption process under different particle sizes is analyzed with infrared thermal imager.The infrared video signals obtained by the experiment are processed with SAT.Then the infrared radiation signals are processed by EMD with Hilbert–Huang and the infrared radiation noise is effectively removed.The research results show that the desorption process,with the change of the temperature,is an endothermic process.The coal absorbs heat when the gas is desorbed and the temperature drops.The coal body temperature drop range is obviously related to coal particle size.The smaller the particle size is,the bigger the temperature drop becomes.The temperature variation curves in the process of coal gas desorption under different particle sizes are fitted,and they comply with the exponential function.The research results lay the theoretical and experimental foundation for non-contact prediction on working face of coal and gas outburst with infrared thermal image technology.

Deep Learning and Time Series-to-Image Encoding for Financial Forecasting

In the last decade,market financial forecasting has attracted high interests amongst the researchers in pattern recognition.Usually,the data used for analysing the market,and then gamble on its future trend,are provided as time series;this aspect,along with the high fluctuation of this kind of data,cuts out the use of very efficient classification tools,very popular in the state of the art,like the well known convolutional neural networks(CNNs)models such as Inception,Res Net,Alex Net,and so on.This forces the researchers to train new tools from scratch.Such operations could be very time consuming.This paper exploits an ensemble of CNNs,trained over Gramian angular fields(GAF)images,generated from time series related to the Standard&Poor's 500 index future;the aim is the prediction of the future trend of the U.S.market.A multi-resolution imaging approach is used to feed each CNN,enabling the analysis of different time intervals for a single observation.A simple trading system based on the ensemble forecaster is used to evaluate the quality of the proposed approach.Our method outperforms the buyand-hold(B&H)strategy in a time frame where the latter provides excellent returns.Both quantitative and qualitative results are provided.

Review of light field technologies

Light fields are vector functions that map the geometry of light rays to the corresponding plenoptic attributes.They describe the holographic information of scenes by representing the amount of light flowing in every direction through every point in space.The physical concept of light fields was first proposed in 1936,and light fields are becoming increasingly important in the field of computer graphics,especially with the fast growth of computing capacity as well as network bandwidth.In this article,light field imaging is reviewed from the following aspects with an emphasis on the achievements of the past five years:(1)depth estimation,(2)content editing,(3)image quality,(4)scene reconstruction and view synthesis,and(5)industrial products because the technologies of lights fields also intersect with industrial applications.State-of-the-art research has focused on light field acquisition,manipulation,and display.In addition,the research has extended from the laboratory to industry.According to these achievements and challenges,in the near future,the applications of light fields could offer more portability,accessibility,compatibility,and ability to visualize the world.

Real-time measurement of welding temperature field and closed loop control of penetration

Colormetric method of images by using two different wavelength images is a new measuring method for welding temperature field on the basis of ordinary colorimetric method, which depends little on the measuring distance, emissivity of body etc. In this paper the real time measuring system and measuring principle of welding temperature field are described, the whole welding temperature field is real time measured, so the temperature distribution at the welding direction and its cross section is obtained, then parameters of thermal cycle. With data from the temperature closed loop control system of the parameters of temperature field is developed and tested. Experimental results prove that it has high measurement speed (time of a field within 0.5 s ) and good dynamic response quality. Weld penetration can be controlled satisfactorily under the variation of welding condition such as welding thickness, welding speed and weldment gap etc.

An easy way to controllably synthesize one-dimensional Sm B_6 topological insulator nanostructures and exploration of their field emission applications

A convenient fabrication technique for samarium hexaboride（SmB6） nanostructures（nanowires and nanopencils） is developed, combining magnetron-sputtering and chemical vapor deposition. Both nanostructures are proven to be single crystals with cubic structure, and they both grow along the [001] direction. Formation of both nanostructures is attributed to the vapor-liquid-solid（VLS） mechanism, and the content of boron vapor is proposed to be the reason for their different morphologies at various evaporation distances. Field emission（FE） measurements show that the maximum current density of both the as-grown nanowires and nanopencils can be several hundred μA/cm^2, and their FN plots deviate only slightly from a straight line. Moreover, we prefer the generalized Schottky-Nordheim（SN） model to comprehend the difference in FE properties between the nanowires and nanopencils. The results reveal that the nonlinearity of FN plots is attributable to the effect of image potential on the FE process, which is almost independent of the morphology of the nanostructures.All the research results suggest that the SmB6 nanostructures would have a more promising future in the FE area if their surface oxide layer was eliminated in advance.

Compressed sensing sparse reconstruction for coherent field imaging

Return signal processing and reconstruction plays a pivotal role in coherent field imaging, having a significant in- fluence on the quality of the reconstructed image. To reduce the required samples and accelerate the sampling process, we propose a genuine sparse reconstruction scheme based on compressed sensing theory. By analyzing the sparsity of the received signal in the Fourier spectrum domain, we accomplish an effective random projection and then reconstruct the return signal from as little as 10% of traditional samples, finally acquiring the target image precisely. The results of the numerical simulations and practical experiments verify the correctness of the proposed method, providing an efficient processing approach for imaging fast-moving targets in the future.

Evaluation of traumatic retinopathy with ultra-wide field imaging under corneal scar or fixed small pupil

AIM： To evaluate the value of ultra-wide field（UWF） imaging in the management of traumatic retinopathy under the condition of corneal scar or fixed small pupil after complicated ocular trauma. METHODS： Twenty-eight patients（28 eyes） with complicated ocular trauma were enrolled in the study from June 2016 to May 2017, including 19 males and 9 females with age ranged from 11 to 64（43.42±12.62）y. All patients were treated with secondary vitrectomy after emergency operation for wound repair of open ocular trauma. Direct ophthalmoscopy and 45-degree fundus photography were taken at each time point of follow up for comparison of findings with UWF images. Routine eye examination including visual acuity, intraocular pressure, slit lamp examination were performed and analyzed as well.RESULTS： Among the 28 traumatized eyes, the positive rate for identification of traumatic retinopathed was 32.1%（9 cases）, 14.9%（5 cases）, and 85.7%（24 cases） with direct ophthalmoscopy, 45-degree fundus photography, and UWF imaging, respectively. The detective rate of UWF imaging under the condition of corneal scar or fixed small pupil was statistically greater than that of 45-degree fundus photography and direct ophthalmoscopy（Bonferroni correction, P〈0.001）. UWF image was obtained in 19 eyes with opaque corneal scar, otherwise their fundus could not be seen by conventional methods. The additional findings of traumatic retinopathies by UWF imaging included periretinal membranes or pre-retinal proliferating strip, retinal holes, hemorrhage in the vitreous or sub-retinal space.CONCLUSION： UWF imaging is superior to traditional fundus photography in the evaluation of traumatic retinopathies under the condition of corneal scar or fixed small pupil after complicated ocular trauma.

Calibration and data restoration of light field modulated imaging spectrometer

A light field modulated imaging spectrometer （LFMIS） can acquire the spatial-spectral datacube of targets of interest or a scene in a single shot. The spectral information of a point target is imaged on the pixels covered by a microlens. The pixels receive spectral information from different spectral filters to the diffraction and misalignments of the optical components. In this paper, we present a linear spectral multiplexing model of the acquired target spectrum. A calibration method is proposed for calibrating the center wavelengths and bandwidths of channels of an LFMIS system based on the liner-variable filter （LVF） and for determining the spectral multiplexing matrix. In order to improve the accuracy of the restored spectral data, we introduce a reconstruction algorithm based on the total least square （TLS） approach. Simulation and experimental results confirm the performance of the spectrum reconstruction algorithm and validate the feasibility of the proposed calibrating scheme.

Wave-electric field coupling imaging diagnostic method for filled subgrade

Thick earth-rock filled embankment of large earthwork volume often occurs during the construction of expressways in mountainous and hilly areas. The compaction quality of earth-rock filled subgrade will directly affect the settlement deformation and stability of the embankment after filled. Therefore, effective evaluation on the compaction quality of the earth-rock filled subgrade is an unsolved critical technical issue to control the construction quality of highway engineering. Based on the wave propagation and electrical resistivity characteristics of the earth and rock fillings, a theoretical model of the compaction quality detection by wave-electric field coupling imaging diagnostic method was established. Then, two filled subgrade models containing cavities and heterogeneous bodies respectively were make separately, and by the wave velocity testing and electrical resistivity testing, the wave-electric field coupling imaging diagnostic method was applied to these two model. The result shows that it is feasible to use the wave testing technique and the electrical resistivity testing technique for a diagnostic test of the subgrade compaction quality. Based on the abnormal areas reflected by the wave velocity imaging and electrical resistivity imaging results, we are able to analyze the scope and site of distress but not able to quantitatively evaluate the subgrade compaction quality. We can accurately qualitatively analyze the subgrade compaction quality based on the wave-electric field coupling calculation model of fill subgrade quality proposed by this paper.

Digital Refocusing: All-in-Focus Image Rendering Based on Holoscopic 3D Camera

This paper presents an innovative method for digital refocusing of different point in space after capturing and also extracts all-in-focus image. The proposed method extracts all-in-focus image using Michelson contrast formula hence, it helps in calculating the coordinates of the 3D object location. With light field integral camera setup the scene to capture the objects precisely positioned in a measurable distance from the camera therefore, it helps in refocusing process to return the original location where the object is focused;else it will be blurred with less contrast. The highest contrast values at different points in space can return the focused points where the objects are initially positioned as a result;all-in-focus image can also be obtained. Detailed experiments are conducted to demonstrate the credibility of proposed method with results.

Measurements of Non-reacting and Reacting Flow Fields of a Liquid Swirl Flame Burner

The understanding of the liquid fuel spray and flow field characteristics inside a combustor is crucial for designing a fuel efficient and low emission device.Characterisation of the flow field of a model gas turbine liquid swirl burner is performed by using a2-D particle imaging velocimetry（PIV）system.The flow field pattern of an axial flow burner with a fixed swirl intensity is compared under confined and unconfined conditions,i.e.,with and without the combustor wall.The effect of temperature on the main swirling air flow is investigated under open and non-reacting conditions.The result shows that axial and radial velocities increase as a result of decreased flow density and increased flow volume.The flow field of the main swirling flow with liquid fuel spray injection is compared to non-spray swirling flow.Introduction of liquid fuel spray changes the swirl air flow field at the burner outlet,where the radial velocity components increase for both open and confined environment.Under reacting condition,the enclosure generates a corner recirculation zone that intensifies the strength of radial velocity.The reverse flow and corner recirculation zone assists in stabilizing the flame by preheating the reactants.The flow field data can be used as validation target for swirl combustion modelling.

Experimental Detection of Depth of Field for a Thermal Light Lensless Ghost Imaging System

We propose optical experiments to study the depth of field for a thermal light lensless ghost imaging system. It is proved that the diaphragm is an important factor to influence the depth of field, and the ghost images of two detected objects with longitudinal distance less than the depth of field can be achieved simultaneously. The longitudinal coherence scale of the thermal light lensless ghost imaging determines the depth of field. Theoretical analysis can well explain the experimental results.

Integrating field images and microclimate data to realize multi-day ahead forecasting of maize crop coverage using CNN-LSTM

Crop coverage(CC)is an important parameter to represent crop growth characteristics,and the ahead forecasting of CC is helpful to track crop growth trends and guide agricultural management decisions.In this study,a novel CNN-LSTM model that combined the advantages of convolutional neural network(CNN)in feature extraction and long short-term memory(LSTM)in time series processing was proposed for multi-day ahead forecasting of maize CC.Considering the influence of climate change on maize growth,five microclimatic factors were combined with historical maize CC estimated from field images as the input variables of the forecasting model.The field experimental data of four observation points for more than three years were used to evaluate the performance of CNN-LSTM at the forecasting horizon of three to seven days ahead and compared the forecasting results to CNN and LSTM.The results demonstrated that CNN-LSTM obtained the lowest RMSE and the highest R2 at all forecasting horizons.Subsequently,the performance of CNN-LSTM under univariate(historical maize CC)and multivariate(historical maize CC+microclimatic factors)input was compared,and the results indicated that additional microclimatic factors were effective in improving the forecasting performance.Furthermore,the 3-day ahead forecasting results of CNN-LSTM in different growth stages of maize were also analyzed,and the results showed that the highest forecasting accuracy was obtained in the seven leaves stage.Therefore,CNN-LSTM can be considered a useful tool to forecast maize CC.

Efficient and robust CNN-LSTM prediction of flame temperature aided light field online tomography

Light field tomography,an optical combustion diagnostic technology,has recently attracted extensive attention due to its easy implementation and non-intrusion.However,the conventional iterative methods are high data throughput,low efficiency and time-consuming,and the existing machine learning models use the radiation spectrum information of the flame to realize the parameter field measurement at the current time.It is still an offline measurement and cannot realize the online prediction of the instantaneous structure of the actual turbulent combustion field.In this work,a novel online prediction model of flame temperature instantaneous structure based on deep convolutional neural network and long short-term memory(CNN-LSTM)is proposed.The method uses the characteristics of local perception,shared weight,and pooling of CNN to extract the threedimensional(3D)features of flame temperature and outgoing radiation images.Moreover,the LSTM is used to comprehensively utilize the ten historical time series information of high dynamic combustion flame to accurately predict 3D temperature at three future moments.A chaotic time-series dataset based on the flame radiation forward model is built to train and validate the performance of the proposed CNN-LSTM model.It is proven that the CNN-LSTM prediction model can successfully learn the evolution pattern of combustion flame and make accurate predictions.

像增强器视场缺陷检测方法研究

在数字式像增强器性能综合检测系统的基础上,研究了像增强器视场缺陷的检测方法。通过数字视频摄像机采集缺陷图像,利用数字图像处理方法确定缺陷性质,缺陷处理算法基于图像灰度特征而不是边缘特征提取。该方法适用于对检测对象有具体量化和定位要求,而不是简单的只有数量和形状特征要求的场合。实验表明,该检测方法能够较准确地检测像增强器的缺陷。