scholarly journals ShadowDeNet: A Moving Target Shadow Detection Network for Video SAR

2022 ◽  
Vol 14 (2) ◽  
pp. 320
Author(s):  
Jinyu Bao ◽  
Xiaoling Zhang ◽  
Tianwen Zhang ◽  
Xiaowo Xu
Keyword(s):  
Feature Extraction ◽  
Adaptive Learning ◽  
Histogram Equalization ◽  
Shadow Detection ◽  
False Alarms ◽  
Moving Target ◽  
The Public ◽  
Motion Speed ◽  
Discrimination Capacity ◽  
Deep Learning Network

Most existing SAR moving target shadow detectors not only tend to generate missed detections because of their limited feature extraction capacity among complex scenes, but also tend to bring about numerous perishing false alarms due to their poor foreground–background discrimination capacity. Therefore, to solve these problems, this paper proposes a novel deep learning network called “ShadowDeNet” for better shadow detection of moving ground targets on video synthetic aperture radar (SAR) images. It utilizes five major tools to guarantee its superior detection performance, i.e., (1) histogram equalization shadow enhancement (HESE) for enhancing shadow saliency to facilitate feature extraction, (2) transformer self-attention mechanism (TSAM) for focusing on regions of interests to suppress clutter interferences, (3) shape deformation adaptive learning (SDAL) for learning moving target deformed shadows to conquer motion speed variations, (4) semantic-guided anchor-adaptive learning (SGAAL) for generating optimized anchors to match shadow location and shape, and (5) online hard-example mining (OHEM) for selecting typical difficult negative samples to improve background discrimination capacity. We conduct extensive ablation studies to confirm the effectiveness of the above each contribution. We perform experiments on the public Sandia National Laboratories (SNL) video SAR data. Experimental results reveal the state-of-the-art performance of ShadowDeNet, with a 66.01% best f1 accuracy, in contrast to the other five competitive methods. Specifically, ShadowDeNet is superior to the experimental baseline Faster R-CNN by a 9.00% f1 accuracy, and superior to the existing first-best model by a 4.96% f1 accuracy. Furthermore, ShadowDeNet merely sacrifices a slight detection speed in an acceptable range.

A new combined transient extraction method coupled with WO3 gas sensors for polluting gases classification

Sensor Review ◽  
2021 ◽  
Vol ahead-of-print (ahead-of-print) ◽  
Author(s):  
Rabeb Faleh ◽  
Sami Gomri ◽  
Khalifa Aguir ◽  
Abdennaceur Kachouri
Keyword(s):  
Feature Extraction ◽  
Gas Sensors ◽  
Classification Accuracy ◽  
Extraction Method ◽  
Support Vector ◽  
Classification Rate ◽  
Content Type ◽  
Feature Extraction Method ◽  
Combined Features ◽  
Discrimination Capacity

Purpose The purpose of this paper is to deal with the classification improvement of pollutant using WO3 gases sensors. To evaluate the discrimination capacity, some experiments were achieved using three gases: ozone, ethanol, acetone and a mixture of ozone and ethanol via four WO3 sensors. Design/methodology/approach To improve the classification accuracy and enhance selectivity, some combined features that were configured through the principal component analysis were used. First, evaluate the discrimination capacity; some experiments were performed using three gases: ozone, ethanol, acetone and a mixture of ozone and ethanol, via four WO3 sensors. To this end, three features that are derivate, integral and the time corresponding to the peak derivate have been extracted from each transient sensor response according to four WO3 gas sensors used. Then these extracted parameters were used in a combined array. Findings The results show that the proposed feature extraction method could extract robust information. The Extreme Learning Machine (ELM) was used to identify the studied gases. In addition, ELM was compared with the Support Vector Machine (SVM). The experimental results prove the superiority of the combined features method in our E-nose application, as this method achieves the highest classification rate of 90% using the ELM and 93.03% using the SVM based on Radial Basis Kernel Function SVM-RBF. Originality/value Combined features have been configured from transient response to improve the classification accuracy. The achieved results show that the proposed feature extraction method could extract robust information. The ELM and SVM were used to identify the studied gases.

scholarly journals The State-of-the-Art Technology of Currency Identification

2017 ◽  
Vol 9 (3) ◽  
pp. 58-72 ◽  
Author(s):  
Guangyu Wang ◽  
Xiaotian Wu ◽  
WeiQi Yan
Keyword(s):  
Feature Extraction ◽  
New Zealand ◽  
Comparative Study ◽  
State Of The Art ◽  
The State ◽  
Classification Algorithms ◽  
Security Issue ◽  
The Public ◽  
Us Dollar ◽  
Chinese Yuan

The security issue of currency has attracted awareness from the public. De-spite the development of applying various anti-counterfeit methods on currency notes, cheaters are able to produce illegal copies and circulate them in market without being detected. By reviewing related work in currency security, the focus of this paper is on conducting a comparative study of feature extraction and classification algorithms of currency notes authentication. We extract various computational features from the dataset consisting of US dollar (USD), Chinese Yuan (CNY) and New Zealand Dollar (NZD) and apply the classification algorithms to currency identification. Our contributions are to find and implement various algorithms from the existing literatures and choose the best approaches for use.

The State-of-the-Art Technology of Currency Identification

2018 ◽  
pp. 252-269
Author(s):  
Guangyu Wang ◽  
Xiaotian Wu ◽  
WeiQi Yan
Keyword(s):  
Feature Extraction ◽  
New Zealand ◽  
Comparative Study ◽  
State Of The Art ◽  
The State ◽  
Classification Algorithms ◽  
Security Issue ◽  
The Public ◽  
Us Dollar ◽  
Chinese Yuan

The security issue of currency has attracted awareness from the public. De-spite the development of applying various anti-counterfeit methods on currency notes, cheaters are able to produce illegal copies and circulate them in market without being detected. By reviewing related work in currency security, the focus of this paper is on conducting a comparative study of feature extraction and classification algorithms of currency notes authentication. We extract various computational features from the dataset consisting of US dollar (USD), Chinese Yuan (CNY) and New Zealand Dollar (NZD) and apply the classification algorithms to currency identification. Our contributions are to find and implement various algorithms from the existing literatures and choose the best approaches for use.

scholarly journals Dog Identification Method Based on Muzzle Pattern Image

2020 ◽  
Vol 10 (24) ◽  
pp. 8994
Author(s):  
Dong-Hwa Jang ◽  
Kyeong-Seok Kwon ◽  
Jung-Kon Kim ◽  
Ka-Young Yang ◽  
Jong-Bok Kim
Keyword(s):  
Feature Extraction ◽  
Radio Frequency Identification ◽  
Hamming Distance ◽  
Histogram Equalization ◽  
Scale Invariant ◽  
Speeded Up Robust Features ◽  
Identification Method ◽  
Approximate Nearest Neighbors ◽  
Frequency Identification ◽  
Scale Invariant Feature

Currently, invasive and external radio frequency identification (RFID) devices and pet tags are widely used for dog identification. However, social problems such as abandoning and losing dogs are constantly increasing. A more effective alternative to the existing identification method is required and the biometrics can be the alternative. This paper proposes an effective dog muzzle recognition method to identify individual dogs. The proposed method consists of preprocessing, feature extraction, matching, and postprocessing. For preprocessing, proposed resize and histogram equalization are used. For feature extraction algorithm, Scale Invariant Feature Transform (SIFT), Speeded Up Robust Features (SURF), Binary Robust Invariant Scaling Keypoints (BRISK) and Oriented FAST, and Rotated BRIEF (ORB) are applied and compared. For matching, Fast Library for Approximate Nearest Neighbors (FLANN) is used for SIFT and SURF, and hamming distance are used for BRISK and ORB. For postprocessing, two techniques to reduce incorrect matches are proposed. The proposed method was evaluated with 55 dog muzzle pattern images acquired from 11 dogs and 990 images augmented by the image deformation (i.e., angle, illumination, noise, affine transform). The best Equal Error Rate (EER) of the proposed method was 0.35%, and ORB was the most appropriate for the dog muzzle pattern recognition.

scholarly journals Multi-Person Pose Estimation using an Orientation and Occlusion Aware Deep Learning Network

Sensors ◽  
2020 ◽  
Vol 20 (6) ◽  
pp. 1593 ◽  
Author(s):  
Yanlei Gu ◽  
Huiyang Zhang ◽  
Shunsuke Kamijo
Keyword(s):  
Deep Learning ◽  
Pose Estimation ◽  
Body Orientation ◽  
Learning Approaches ◽  
The Public ◽  
Learning Network ◽  
Proposed Model ◽  
Body Boundary ◽  
Public Dataset ◽  
Deep Learning Network

Image based human behavior and activity understanding has been a hot topic in the field of computer vision and multimedia. As an important part, skeleton estimation, which is also called pose estimation, has attracted lots of interests. For pose estimation, most of the deep learning approaches mainly focus on the joint feature. However, the joint feature is not sufficient, especially when the image includes multi-person and the pose is occluded or not fully visible. This paper proposes a novel multi-task framework for the multi-person pose estimation. The proposed framework is developed based on Mask Region-based Convolutional Neural Networks (R-CNN) and extended to integrate the joint feature, body boundary, body orientation and occlusion condition together. In order to further improve the performance of the multi-person pose estimation, this paper proposes to organize the different information in serial multi-task models instead of the widely used parallel multi-task network. The proposed models are trained on the public dataset Common Objects in Context (COCO), which is further augmented by ground truths of body orientation and mutual-occlusion mask. Experiments demonstrate the performance of the proposed method for multi-person pose estimation and body orientation estimation. The proposed method can detect 84.6% of the Percentage of Correct Keypoints (PCK) and has an 83.7% Correct Detection Rate (CDR). Comparisons further illustrate the proposed model can reduce the over-detection compared with other methods.

VISUAL TARGET TRACKING, EXTRACTION AND RECOGNITION FROM A SEQUENCE OF IMAGES USING THE LG GRAPH APPROACH

2002 ◽  
Vol 11 (04) ◽  
pp. 513-529 ◽  
Author(s):  
NIKOLAOS G. BOURBAKIS
Keyword(s):  
Feature Extraction ◽  
Edge Detection ◽  
Target Tracking ◽  
Visual Target ◽  
Region Growing ◽  
Geometric Approach ◽  
Moving Target ◽  
Extraction Region ◽  
Fuzzy Like ◽  
Visual Target Tracking

This paper presents a methodology for visually tracking, extracting and recognizing targets from a sequence of images (video). The methodology is based on the local-global (LG) graph as a combination of algorithms, such as fuzzy-like segmentation, edge detection, thinning, region growing, fractals, feature extraction, region-graph with attributes, etc., appropriately used for tracking, extracting and recognizing targets under various conditions, such as moving target - still camera, still camera - moving target, moving target - moving camera. The main contribution of this paper is the real-time combination of algorithms that provides a human-like feedback geometric approach of processing low resolution information in a sequence of consecutive images. Simulated results of the metholodology are presented for synthetic and real images.

Sign in / Sign up

Export Citation Format

Share Document