For detecting rice disease early and accurately, we presented an application of image processing techniques andSupport Vector Machine (SVM) for diseases.Rice spots were segmented their shape texture features extracted. The SVM method was employed to classify bacterial leaf blight, sheath blight blast. results showed that could effectively detect these accuracy 97.2%.

Non-destructive plant growth measurement is essential for and health research. As a 3D sensor, Kinect v2 has huge potentials in agriculture applications, benefited from its low price strong robustness. The paper proposes Kinect-based automatic system non-destructive of leafy vegetables. used turntable to acquire multi-view point clouds the measured plant. Then series suitable algorithms were applied obtain fine reconstruction plant, while measuring key parameters including relative/absolute...

In this paper, a high resolution and response speed interrogation method based on reflective-matched Fiber Bragg Grating (FBG) scheme is investigated in detail. The nonlinear problem of the FBG sensing solved by establishing optimizing mathematical model. A mechanical adjustment to optimize tuning central wavelength reference improve stability anti-temperature perturbation performance investigated. To satisfy measurement requirements optical electric signal processing, well- designed...

The calculation of binary iris codes from feature values (e.g., the result Gabor transform) is a key step in recognition systems. Traditional binarization method based on sign has achieved very promising performance. However, currently, little research focuses deeper insight into this to produce codes. In paper, we illustrate code perspective optimization. We demonstrate that traditional solution an optimization problem, which minimizes distance between and Furthermore, show more effective...

An optimized medical image compression algorithm based on wavelet transform and improved vector quantization is introduced. The goal of the proposed method to maintain diagnostic-related information at a high ratio. Wavelet transformation was first applied image. For lowest-frequency subband coefficients, lossless exploited; for each high-frequency subbands, an with variable block size implemented. In novel method, local fractal dimension (LFD) used analyze complexity subband. Then optimal...

Through their breath, humans exhale hundreds of volatile organic compounds (VOCs) that can reveal pathologies, including many types cancer at early stages. Gas chromatography-mass spectrometry (GC-MS) is an analytical method used to separate and detect in the mixture contained breath samples. The identification VOCs based on recognition specific ion patterns GC-MS data, which requires labour-intensive time-consuming preprocessing analysis by domain experts. This paper explores original idea...

A fiber probe with a novel structure of twin Bragg grating sensitive to both contacts along x-axis and z-axis is proposed for measurement microholes high aspect ratio. The main conundrums during the dimensional ratio, such as shadowing effect, limited probing space, requirement multidimensional tactile sense are well solved using probe. Sensing principle decoupling algorithm has been verified through experiments. Experimental results indicate that influence manufacture defects misalignment...

The spectral purity of fiber lasers has become a critical issue in both optical sensing and communication fields. As result ultra-narrow intrinsic linewidth, stimulated thermal Rayleigh scattering (STRS) presented special potential to compress the linewidth lasers. To suppress Brillouin (SBS), most dominant disturbance for STRS fibers, semi-quantitative estimation been established illuminate mechanism suppressing SBS periodic tapered fiber, it agrees with experimental results. Finally,...

This report demonstrates our solution for the Open Images 2018 Challenge. Based on detailed analysis Datasets (OID), it is found that there are four typical features: large-scale, hierarchical tag system, severe annotation incompleteness and data imbalance. Considering these characteristics, an amount of strategies employed, including SNIPER, soft sampling, class-aware sampling (CAS), non-maximum suppression (HNMS) so on. In virtue effective strategies, further using powerful SENet154 armed...

The scarcity of high-resolution urban digital elevation model (DEM) datasets, particularly in certain developing countries, has posed a challenge for many water-related applications such as flood risk management. A solution to address this is develop effective approaches reconstruct DEMs from their low-resolution equivalents that are more widely available. However, the current DEM reconstruction mainly focus on natural topography. Few attempts have been made topography, which typically an...

This paper considers the problem of direction-of-arrival (DOA) estimation coherent signals on passive coprime arrays, where we resort to fourth-order cumulants received signal explore more information. A cumulant matrix (FCM) is introduced for array. The special structure FCM combined with array configuration resolve signals. Since each sparse a uniform, series overlapping identical subarrays can be extracted. Using this property, propose generalized spatial smoothing scheme applied FCM....

Snow and ice on asphalt pavement surface can significantly affect traffic safety damage pavements. To reduce its effect, a kind of anti-icing emulsified binder (AEAB) is developed by adding certain amount fillers to the modified binder. The performance AEAB coating are evaluated. results show that: basic properties meet requirement road maintenance, has excellent bonding property. It facilitate ice-melting adhesion force between pavements ice, be reduced about 50%. keep an state with...

In this paper, we address the problem of iris recognition under less constrained environment. We propose a novel weight map for matching stage to improve robustness noise and degradations in The proposed is class specific considering both bit stability discriminability codes. It combination map. focuses on intra-class stability, aiming matching. assigns more bits that are highly consistent with their noiseless estimations which sought via low rank approximation. models inter-class...

License Plate Recognition (LPR) is a very important research topic in computer vision of ITS. plate location the key step LPR. Though numerous techniques have been developed, most approaches work only under restricted conditions such as fixed illumination, limited vehicle license plates,and simple backgrounds. This paper attempts to use AdaBoost algorithm build up classifiers based on various features. Combining using different features, we obtain cascade classifier. Then classifier which...

Breast cancer continues to be an attentional public health problem all over the world. Because of being indistinguishable from surrounding normal tissues, automated mass detection is challenging in computer-aided (CAD) system. This paper presents a novel system for digital mammograms which integrates visual saliency model with deep learning techniques. After applied pre-processed mammogram, sliding window strategy adopted scan breast area segmented whole mammogram. For each subimage produced...

Purpose Soft robotics is a burgeoning field owing to its high adaptability and safety in human–machine interaction unstructured environments. However, the feedback control of soft actuators with flexible sensors still challenge. Design/methodology/approach To address this issue, study proposes an optical fibre-based sensing membrane for posture measurement pneumatic bending actuators. The major contribution development sensitivity repeatability characteristics were analysed. relationship...

Iris recognition systems working in less constrained environments with the subject at-a-distance and on-the-move suffer from noise degradations iris captures. These significantly deteriorate performance. In this paper, we propose a novel signal-level information fusion method to mitigate influence of for systems. The proposed is based on low rank approximation (LRA). Given multiple noisy captures same eye, assume that: 1) potential noiseless images lie subspace 2) spatially sparse. Based...

Focusing on the ultra-precision dimensional measurement of parts with micro-scale dimensions and high aspect ratios, a two-dimensional double fiber probe single Bragg grating (DS-FBG probe) is investigated in detail this paper. The theoretical analysis sensing principle verified by spectrum simulations DS-FBG modified transfer matrix method using strain distribution within probe. fabrication process physical capillary-driven self-assembly fibers UV adhesive low viscosity are demonstrated....

We report the performances of cancer cell classification by using supervised and unsupervised learning techniques. A single hidden layer feedforward NN with error back-propagation training is adopted for learning, c-means clustering methods, fuzzy nonfuzzy, are used learning. Network configurations various activation functions, namely sigmoid, sinusoid gaussian, studied. set features, including size, average intensity, texture, shape factor pgDNA selected as input network. These in...

Recognition of handwritten mathematical expressions (HMEs) has become a cutting edge research topic recently, as there are increasingly needs for pen-inputting applications. In this paper, we presented novel framework to analyse HME layout and semantic information. This includes three steps, namely symbol segmentation, recognition relationship analysis. For decomposition on strokes is operated, then dynamic programming adopted find the paths corresponding best segmentation manner reduce...