Normally functioning and complete printed circuit board (PCB) can ensure the safety reliability of electronic equipment. PCB defect detection is extremely important in field industrial inspection. For traditional methods inspection, such as contact detection, are likely to damage surface have high rate erroneous detection. In recent years, through image processing machine learning gradually been put into use. However, inspection still an challenging task due small defects complex background....

This paper adresses the important problem of speech detection. It describes implementation 3 detection methods and compares their performance under different signal-to-noise ratio (SNR) stationarity conditions. The method that dynamically adjusts its thresholds is found to be most reliable, even very adverse recording Yet it low complexity has a moderate processing delay.

The rapid development of the license plate recognition technology has made great progress for its widespread uses in intelligent transportation system (ITS). This paper proposed a novel detection and character algorithm based on combined feature extraction model BPNN (Backpropagation Neural Network) which is adaptable weak illumination complicated backgrounds. Firstly, preprocessing first used to strengthen contrast ratio original car image. Secondly, candidate regions are checked verify...

The rapid development of unmanned aerial vehicles (UAVs) has made great progress for its widespread uses in military and civilian applications recent years. On‐board integrated navigation sensors are essential UAV flight control systems that they must operate with robustness reliability. To achieve this, timely effectively anomaly detection capabilities the estimated status from required to ensure safety. Extraction information real‐time designing a robust reliable algorithm major issues...

Indoor localization has long been a challenging task due to the complexity and dynamism of indoor environments. This paper proposes ULG-SLAM, novel unsupervised learning geometric-based visual SLAM algorithm for robot localizability estimation improve accuracy robustness SLAM. Firstly, dynamic feature filtering based on moving consistency checks is developed eliminate features objects. Secondly, an improved line extraction LSD proposed optimize effect geometric extraction. Thirdly, are used...

Some noisy nonlinear systems could be exploited to operate reliable logic operation in an optimal window of noise intensity, which is termed as logical stochastic resonance (LSR). We investigated the LSR phenomenon bistable when internal and external are correlated. The effect evaluated by success probability obtained desired output with various combinations inputs. It shown that or-nor, and-nand, Latch operations still can reliably correlated noise. A positive correlation strength tends...

Abstract Considering a class of quadrotor with variable mass load and external disturbance, novel control strategy based on adaptive non‐singular fast terminal sliding mode is proposed to in this article. Like disturbances, internal disturbance caused by change will further affect the stability system reduce reliability system. Firstly, dynamic modeling carried out Euler Newton formulation. Then, controller integral backstepping designed overcome some disturbances. Meanwhile, an estimation...

The majority of planning algorithms used are based on the occupancy grid maps, but in complicated situations, maps have a significant search overhead. This paper proposed path planner visibility graph (v-graph) for mobile robot that uses sparse methods to speed up and simplify construction v-graph. Firstly, complementary framework is designed reduce updating iteration costs during data collection process each frame. Secondly, filter approach edge length number vertices obstacle contour...

In this paper we present a family of nonlinear spectral estimators for noise reduction which are approximated and implemented by multilayer perceptron neural network. The approximations the true minimum mean square error estimator in logarithmic or related perceptual domain. Training data networks is generated from relevant statistical speech models. One single network all frequency channels. Parameters describing both distribution estimated on line provided as extra inputs to net. Including...

The detection of driving events could be useful for reducing accidents, fleet management and insurance premiums etc. Currently, top the range vehicles large fleets employ expensive driver monitoring systems. However, most drivers do not have access to such required platform would deliver performance while also being affordable accessible. A candidate with considerable promise is smartphone sensors built-in that exploited events. date it has been possible achieve correct, missed false rates...

Abstract Motivation The third generation PacBio long reads have greatly facilitated sequencing projects with very large read lengths, but they contain about 15% errors and need error correction. For the only, it is challenging to make correction fast speed, also correct a sufficient amount of bases, i.e. achieve high-throughput self-correction. MECAT currently among fastest self-correction algorithms, its throughput relatively small (Xiao et al., 2017). Results Here, we introduce FLAS,...

Abstract The canonical description of structures lies in two aspects: 1) basic structural elements and 2) arrangement pattern between those elements. This tidy is very successful facilitates the development structure physics tremendously. Through this, people become capable classifying, comparing, analyzing an extremely wide range structures, including crystals, quasi-crystals, liquid semi-crystalline materials, so on. However, it gradually realizes that many novel materials devices exhibit...

<sec> <title>UNSTRUCTURED</title> Background: Tongue characteristics, widely utilized in traditional Chinese medicine for health assessment, have been shown to correlate with specific respiratory infections. With the ongoing global spread of Human adenoviruses (HAdVs), COVID-19, and other seasonal viruses, this study aims enhance convenience cost-effectiveness infection diagnoses by developing prediction models based on tongue characteristics. Method: This deep learning extract features from...

We are concerned with using user-tagged images to learn proper hashing functions for image retrieval. The benefits two-fold: (1) we could obtain abundant training data deep models; (2) tagging possesses richer semantic information which help better characterize similarity relationships between images. However, suffers from noises, vagueness and incompleteness. Different previous unsupervised or supervised learning, propose a novel weakly-supervised framework consists of two stages:...

One of the important factors leading to traffic accidents is abnormal driving behavior drivers. Early detection behaviors can effectively reduce occurrence accidents. At present, most mainstream methods are based on data a single moment, which separates continuity behavior. In this article, recognition algorithm Serial-Feature Network (SF-Net) and smart phone inertial sensor proposed, fully considers events uses adjacent multi time identify status. The used in article collected from GPS...

With the development of global navigation satellite system (GNSS), GNSS/inertial (INS) integrated offers users better positioning or performance. This paper proposes an adaptive robust ultra‐tightly coupled GNSS/INS based on a novel vector tracking strategy for combining both (GPS) L1 and BeiDou B1 signals’ together. The inherent mechanism approach has been analysed to describe relationship between replica signals user's dynamic state. Then, filter is used gain accurate estimates vehicle...

In this paper, a novel adaptive dynamic sliding mode control(ADSMC) methodology is developed for space manipulator control systems with external disturbance. The proposed ADSMC approach can make the stable and accurately track desired signals in presence of disturbances. Firstly, simplified model introduced. Then, mode(ADSM) controller given, which proved to be able guarantee high tracking accuracy via effectively estimating disturbance boundary. Finally, experimental results show that paper...

In order to improve the accuracy of visual SLAM algorithms in a dynamic scene, instance segmentation is widely used eliminate feature points. However, existing technology has low accuracy, especially for contour object, and amount calculation large, limiting speed based on segmentation. Therefore, this paper proposes optimization hybrid dilated convolutional neural network (CO-HDC) algorithm, which can perform lightweight basis improving Firstly, (HDC) increase receptive field, defined as...

Mobile robots’ efficient path planning has long been a challenging task due to the complexity and dynamism of environments. If an occupancy grid map is used in planning, number grids determined by resolution size actual environment. Excessively high increases traversed nodes thus prolongs time. To address this challenge, paper proposes algorithm based on laser SLAM optimized visibility graph for mobile robots, which achieves faster computation shortest using graph. Firstly, acquire...