scholarly journals Dim and Small Targets Detection in Sequence Images Based on Spatiotemporal Motion Characteristics

2020 ◽  
Vol 2020 ◽  
pp. 1-19
Author(s):  
Fan Xiangsuo ◽  
Hongwei Guo ◽  
Xu Zhiyong ◽  
Biao Li
Keyword(s):  
Detection Algorithm ◽  
Spatial Domain ◽  
Mean Value ◽  
Local Motion ◽  
Detection Rates ◽  
Motion Characteristics ◽  
Small Targets ◽  
The Time Domain ◽  
Diffusion Function ◽  
Sequential Images

In order to effectively enhance the low detection rates of dim and small targets caused by dynamic backgrounds, this paper proposes a detection algorithm for dim and small targets in sequence images based on spatiotemporal motion characteristics. With regard to the spatial domain, this paper proposes an improved anisotropic background filtering algorithm that makes full use of the gradient differences between the target and the background pixels in the eight directions of the spatial domain and selects the mean value of the three directions with the lowest diffusion function in the eight directions as the differential filter to obtain a differential image. Then, the paper proposes a directional energy correlation enhancement algorithm in the time domain. Finally, based on the above preprocessing operations, we construct a dim and small targets detection algorithm in sequence images with local motion characteristics, which achieves target detection by determining the number of occurrences of the target, the number of displacements, and the total cumulative area in these sequential images. Experiments show that the proposed detection algorithm in this paper can effectively improve the detection of dim and small targets in dynamic scenes.

scholarly journals Infrared Dim and Small Targets Detection Method Based on Local Energy Center of Sequential Image

2017 ◽  
Vol 2017 ◽  
pp. 1-16 ◽  
Author(s):  
Xiangsuo Fan ◽  
Zhiyong Xu ◽  
Jianlin Zhang ◽  
Yongmei Huang ◽  
Zhenming Peng
Keyword(s):  
Detection Method ◽  
Detection Algorithm ◽  
Local Energy ◽  
Signal Noise ◽  
Sequential Image ◽  
Energy Center ◽  
Overall Performance ◽  
Noise Ratio ◽  
Motion Characteristics ◽  
Small Targets

In order to detect infrared (IR) dim and small targets in a strong clutter background, a method based on local energy center of sequential image is proposed. This paper began by using improved anisotropy for background prediction (IABP), followed by target enhancement by improved high-order cumulates (HOC). Finally, on the basis of image preprocessing, the paper constructs a sequential image energy center detection algorithm that integrates the neighborhood, continuity, area, and energy and other motion characteristics of the target. Experiments showed that the improved anisotropic background predication could be loyal to the true background of the original image to the maximum extent, presenting a superior overall performance to other background prediction methods; the improved HOC significantly increased the signal-noise ratio of images; when the signal-noise ratio (SNR) is lower than 2.5 dB, the proposed method could effectively eliminate noise and detect targets.

scholarly journals Head Detection Based on DR Feature Extraction Network and Mixed Dilated Convolution Module

Electronics ◽  
2021 ◽  
Vol 10 (13) ◽  
pp. 1565
Author(s):  
Junwen Liu ◽  
Yongjun Zhang ◽  
Jianbin Xie ◽  
Yan Wei ◽  
Zewei Wang ◽  
...  
Keyword(s):  
Neural Network ◽  
Feature Extraction ◽  
Transmission Rate ◽  
Pedestrian Detection ◽  
Human Head ◽  
Detection Rates ◽  
Translational Invariance ◽  
Dilated Convolution ◽  
Head Detection ◽  
Small Targets

Pedestrian detection for complex scenes suffers from pedestrian occlusion issues, such as occlusions between pedestrians. As well-known, compared with the variability of the human body, the shape of a human head and their shoulders changes minimally and has high stability. Therefore, head detection is an important research area in the field of pedestrian detection. The translational invariance of neural network enables us to design a deep convolutional neural network, which means that, even if the appearance and location of the target changes, it can still be recognized effectively. However, the problems of scale invariance and high miss detection rates for small targets still exist. In this paper, a feature extraction network DR-Net based on Darknet-53 is proposed to improve the information transmission rate between convolutional layers and to extract more semantic information. In addition, the MDC (mixed dilated convolution) with different sampling rates of dilated convolution is embedded to improve the detection rate of small targets. We evaluated our method on three publicly available datasets and achieved excellent results. The AP (Average Precision) value on the Brainwash dataset, HollywoodHeads dataset, and SCUT-HEAD dataset reached 92.1%, 84.8%, and 90% respectively.

Study of Snubbing Mechanism Using Finite Element Method

2014 ◽  
Author(s):  
Jingming Chen ◽  
Paolo Pennacchi ◽  
Dongxiang Jiang ◽  
Steven Chatterton
Keyword(s):  
Finite Element Method ◽  
Finite Element ◽  
Forced Vibration ◽  
Excitation Frequency ◽  
Relative Displacement ◽  
Rotor Blades ◽  
Motion Pattern ◽  
Motion Characteristics ◽  
The Time Domain ◽  
Element Method

In the rotating machineries, large vibrations of a blade would result in fatigue crack, which is a great threaten to the safety. Therefore, it is of great importance to reduce the blade vibrations. Snubbing technique is a possible solution to this problem. A tiny gap is left between the shrouds of adjacent blades. While the forced vibration makes the relative displacement between two neighboring blades exceed the gap, the contact happens at the contact face of the shrouds, accompanied with friction and energy dissipation, which restricts the vibration. In this paper, a simplified model for a set of rotor blades is established, by using finite element method. The contact between the adjacent shrouds is considered. In this way, snubbing phenomenon can occur under forced vibration. Based on the model, modal analysis has been conducted. The 8x rev. frequency has been chosen as the excitation frequency. Under a certain amplitude of sine excitation, the circumferential vibration of the blades has been simulated. The vibration has been analyzed in the time domain. As expected, the blade motion is divided into four different states in one period. They are: non-contact, rebounding, sticky and escaping state. The four states had different mechanical and motion characteristics. The motion pattern for the set of blades has been also analyzed and the wave spreading along the bladerow has been described. Because of the snubbing mechanism, the waveform was distorted into serrated shape.

scholarly journals Pedestrian detection algorithm based on improved muti-scale feature fusion

2021 ◽  
Vol 2078 (1) ◽  
pp. 012008
Author(s):  
Hui Liu ◽  
Keyang Cheng
Keyword(s):  
Clustering Algorithm ◽  
Feature Fusion ◽  
Pedestrian Detection ◽  
Detection Algorithm ◽  
Data Sets ◽  
False Detection ◽  
Scale Feature ◽  
Multi Scale ◽  
Dilated Convolution ◽  
Small Targets

Abstract Aiming at the problem of false detection and missed detection of small targets and occluded targets in the process of pedestrian detection, a pedestrian detection algorithm based on improved multi-scale feature fusion is proposed. First, for the YOLOv4 multi-scale feature fusion module PANet, which does not consider the interaction relationship between scales, PANet is improved to reduce the semantic gap between scales, and the attention mechanism is introduced to learn the importance of different layers to strengthen feature fusion; then, dilated convolution is introduced. Dilated convolution reduces the problem of information loss during the downsampling process; finally, the K-means clustering algorithm is used to redesign the anchor box and modify the loss function to detect a single category. The experimental results show that the improved pedestrian detection algorithm in the INRIA and WiderPerson data sets under different congestion conditions, the AP reaches 96.83% and 59.67%, respectively. Compared with the pedestrian detection results of the YOLOv4 model, the algorithm improves by 2.41% and 1.03%, respectively. The problem of false detection and missed detection of small targets and occlusion has been significantly improved.

Wavenumber Spectra of High Magnitude Wall Pressure Events in a Numerically Simulated Turbulent Boundary Layer

1997 ◽  
Vol 119 (2) ◽  
pp. 281-288
Author(s):  
B. M. Abraham ◽  
W. L. Keith
Keyword(s):  
Boundary Layer ◽  
Turbulent Boundary Layer ◽  
Fourier Transforms ◽  
Experimental Studies ◽  
Detection Algorithm ◽  
Wall Pressure ◽  
Negative Events ◽  
Discrete Fourier Transforms ◽  
High Magnitude ◽  
The Time Domain

A method for conditionally sampling the spatial field of the wall pressure beneath a turbulent boundary layer in order to search for high magnitude events and calculate the corresponding wavenumber spectrum is presented. The high magnitude events are found using a simple peak detection algorithm at a fixed instant in time and the wavenumber spectra are calculated using discrete Fourier transforms. The frequency of occurrence for high magnitude positive events is found to be approximately the same as for high magnitude negative events. The contribution of the high magnitude events to the rms wall pressure for various trigger levels is calculated and compared with results from similar experimental studies performed in the time domain. The high magnitude events are shown to occur infrequently and to contribute significantly to the rms wall pressure. Wavenumber spectra from the high magnitude positive and negative events are calculated and compared with the unconditionally sampled spectra. The high magnitude events contain energy focused around a particular stream-wise wavenumber and have high broadband spectral levels.

scholarly journals The MEMO project: Combining all microlensing surveys to search for intermediate-mass Galactic black holes

2018 ◽  
Vol 618 ◽  
pp. L4 ◽  
Author(s):  
A. Mirhosseini ◽  
M. Moniez
Keyword(s):  
Black Holes ◽  
Black Hole ◽  
Milky Way ◽  
Detection Efficiency ◽  
Large Magellanic Cloud ◽  
Detection Algorithm ◽  
Intermediate Mass ◽  
Detection Rates ◽  
Galactic Black Hole ◽  
The Individual

Aims. The microlensing surveys MACHO, EROS, MOA and OGLE (hereafter called MEMO) have searched for microlensing toward the Large Magellanic Cloud for a cumulated duration of 27 years. We study the potential of joining these databases to search for very massive objects, that produce microlensing events with a duration of several years. Methods. We identified the overlaps between the different catalogs and compiled their time coverage to identify common regions where a joint microlensing detection algorithm can operate. We extrapolated a conservative global microlensing detection efficiency based on simple hypotheses, and estimated detection rates for multi-year duration events. Results. Compared with the individual survey searches, we show that a combined search for long timescale microlensing should detect about ten more events caused by 100 M⊙ black holes if these objects have a major contribution to the Milky Way halo. Conclusions. Assuming that a common analysis is feasible, meaning that the difficulties that arise from using different passbands can be overcome, we show that the sensitivity of such an analysis might enable us to quantify the Galactic black hole component.

INFRARED SMALL TARGET DETECTION ALGORITHM BASED ON FEATURE SALIENCE AND MULTI-FEATURES FUSION

2011 ◽  
Vol 25 (02) ◽  
pp. 299-308 ◽  
Author(s):  
ZHEN-XUE CHEN ◽  
CHENG-YUN LIU ◽  
FA-LIANG CHANG
Keyword(s):  
Target Detection ◽  
Detection Algorithm ◽  
Probability Of Detection ◽  
Absolute Difference ◽  
Small Target ◽  
Probability Of Error ◽  
Small Target Detection ◽  
Low Snr ◽  
Complex Background ◽  
Small Targets

It is an important and challenging problem to detect small targets in clutter scene and low SNR (Signal Noise Ratio) in infrared (IR) images. In order to solve this problem, a method based on feature salience is proposed for automatic detection of targets in complex background. Firstly, in this paper, the method utilizes the average absolute difference maximum (AADM) as the dissimilarity measurement between targets and background region to enhance targets. Secondly, minimum probability of error was used to build the model of feature salience. Finally, by computing the realistic degree of features, this method solves the problem of multi-feather fusion. Experimental results show that the algorithm proposed shows better performance with respect to the probability of detection. It is an effective and valuable small target detection algorithm under a complex background.

scholarly journals Pitch Estimation using Time Domain Algorithm

2019 ◽  
Vol 8 (11) ◽  
pp. 1391-1394
Keyword(s):  
Time Domain ◽  
Vocal Folds ◽  
Detection Algorithm ◽  
Pitch Detection ◽  
Pitch Estimation ◽  
The Time Domain ◽  
Short Time ◽  
The Voice ◽  
Magnitude Difference ◽  
Average Magnitude Difference Function

Speech is classified into voice, unvoiced and silence. The voice speech is the periodic vibration of vocal folds. Background noise affects the speech signals. In many speech applications calculation of pitch plays a major role. The paper proposes a pitch detection algorithm based on the short-time average magnitude difference function (AMDF) and the short-term autocorrelation function (ACF). Detecting the Pitch within the speech signal is important in most of all the speech related applications. Detection of Pitch is useful in identification of speaker. One solution to get detect with the pitch is by using the time domain algorithms. This paper gives idea about estimation and detection of pitch in time domain algorithm for different voice samples

scholarly journals Detection method based on improved faster R-CNN for pin defect in transmission lines

2021 ◽  
Vol 300 ◽  
pp. 01011
Author(s):  
Jun Wu ◽  
Sheng Cheng ◽  
Shangzhi Pan ◽  
Wei Xin ◽  
Liangjun Bai ◽  
...  
Keyword(s):  
Transmission Line ◽  
Defect Detection ◽  
Transmission Lines ◽  
False Positive Rate ◽  
Detection Algorithm ◽  
Detection Model ◽  
Positive Rate ◽  
Aerial Vehicle ◽  
The Stability ◽  
Small Targets

Defects such as insulator, pins, and counterweight in highvoltage transmission lines affect the stability of the power system. The small targets such as pins in the unmanned aerial vehicle (UAV) inspection images of transmission lines occupy a small proportion in the images and the characteristic representations are poor which results a low defect detection rate and a high false positive rate. This paper proposed a transmission line pin defect detection algorithm based on improved Faster R-CNN. First, the pre-training weights with higher matching degree are obtained based on transfer learning. And it is applied to construct defect detection model. Then, the regional proposal network is used to extract features in the model. The results of defect detection are obtained by regression calculation and classification of regional characteristics. The experimental results show that the accuracy of the pin defect detection of the transmission line reaches 81.25%

Sign in / Sign up

Export Citation Format

Share Document