COMPARATIVE ANALYSIS OF HOG WITH LINEAR SVM AND SSD ALGORITHMS FOR REAL-TIME PEDESTRIAN DETECTION

2019 ◽  
Author(s):  
Kalpesh Kuber
Author(s):  
V. Нolovan ◽  
V. Gerasimov ◽  
А. Нolovan ◽  
N. Maslich

Fighting in the Donbas, which has been going on for more than five years, shows that a skillful counter-battery fight is an important factor in achieving success in wars of this kind. Especially in conditions where for the known reasons the use of combat aviation is minimized. With the development of technical warfare, the task of servicing the counter-battery fight began to rely on radar stations (radar) to reconnaissance the positions of artillery, which in modern terms are called counter-battery radar. The principle of counter-battery radar is based on the detection of a target (artillery shell, mortar mine or rocket) in flight at an earlier stage and making several measurements of the coordinates of the current position of the ammunition. According to these data, the trajectory of the projectile's flight is calculated and, on the basis of its prolongation and extrapolation of measurements, the probable coordinates of the artillery, as well as the places of ammunition falling, are determined. In addition, the technical capabilities of radars of this class allow you to recognize the types and caliber of artillery systems, as well as to adjust the fire of your artillery. The main advantages of these radars are:  mobility (transportability);  inspection of large tracts of terrain over long distances;  the ability to obtain target's data in near real-time;  independence from time of day and weather conditions;  relatively high fighting efficiency. The purpose of the article is to determine the leading role and place of the counter-battery radar among other artillery instrumental reconnaissance tools, to compare the combat capabilities of modern counter-battery radars, armed with Ukrainian troops and some leading countries (USA, China, Russia), and are being developed and tested in Ukraine. The method of achieving this goal is a comparative analysis of the features of construction and combat capabilities of modern models of counter-battery radar in Ukraine and in other countries. As a result of the conducted analysis, the directions of further improvement of the radar armament, increasing the capabilities of existing and promising counter-battery radar samples were determined.


2018 ◽  
Vol 1069 ◽  
pp. 012107
Author(s):  
Jijun Yang ◽  
Yingdong Ma ◽  
Zhibin Zhang

2021 ◽  
Vol 2021 ◽  
pp. 1-7
Author(s):  
Zhaoli Wu ◽  
Xin Wang ◽  
Chao Chen

Due to the limitation of energy consumption and power consumption, the embedded platform cannot meet the real-time requirements of the far-infrared image pedestrian detection algorithm. To solve this problem, this paper proposes a new real-time infrared pedestrian detection algorithm (RepVGG-YOLOv4, Rep-YOLO), which uses RepVGG to reconstruct the YOLOv4 backbone network, reduces the amount of model parameters and calculations, and improves the speed of target detection; using space spatial pyramid pooling (SPP) obtains different receptive field information to improve the accuracy of model detection; using the channel pruning compression method reduces redundant parameters, model size, and computational complexity. The experimental results show that compared with the YOLOv4 target detection algorithm, the Rep-YOLO algorithm reduces the model volume by 90%, the floating-point calculation is reduced by 93.4%, the reasoning speed is increased by 4 times, and the model detection accuracy after compression reaches 93.25%.


2020 ◽  
Vol 8 (2) ◽  
pp. 69-76 ◽  
Author(s):  
Gulbir Singh ◽  
Vivek Bhatnagar ◽  
Rajeev Gupta ◽  
Gautam Kumar

Purpose of Study: The purpose of the study is to carry out the comparison of traditional learning and e-learning with reference to university education. Methodology: A comparative analysis method that has been selected to fulfill the purpose of the study. The research data has been collected by various sources internet, previous research studies published in the Journal, universities which offer different technical and nontechnical programs .and further analyzed by in-depth understanding approach. Result: The outcome of this study will show the effect of e-learning in the current era. As we have compared e-learning with traditional learning and the result shows us that e-learning has filled the gap between the universities and the real-life industries' demands. Main finding: The result of this study showed that e-learning is the need of the present era in order to fulfill the gap between the universities and the real-life industries' demands. Additionally, e-learning based courses will have more impact and provide more skill and exposure to students as compared to traditional learning mode. Implications/Applications: This article can help the different universities and students to introduce more e-learning courses in their curriculum and they can fill the real-time industries' demands. Novelty/originality of the study: Our research can ensure that e-learning has a huge impact on our students and can help to increase the skills and exposure of them according to the current demands.


Author(s):  
Maryna Sapachova

ObjectiveThe performance of comparative analysis of sensitivity and resultsof detection of avian influenza virus by real time polymerase chainreaction (PCR-RT) and loop-mediated isothermal amplification of thenucleic acids (LAMP) was the main goal of the study.IntroductionAs part of this surveillance study for Avian Influenza both activeand passive surveillance samples were tested using PCR and alsoutilized to validate the LAMP method. Active surveillance samplesinclude pathological material and tracheal and cloacal swabs fromill poultry, which were subsequently assessed for avian influenzaduring diagnosis, and birds collected by hunters. Passive surveillanceincluded environmental samples such as sand and bird faeces.Active surveillance samples were taken mostly from poultry farmsacross Ukraine, where infected birds are required to be diagnosedby State Scientific Research Institute of Laboratory Diagnosticsand Veterinary Sanitary Expertise (SSRILDVSE) by Ukraine Law.Passive surveillance samples were taken primarily during the annualbird migration season. Development of simple, sensitive, and cheapmethods for diagnostics of avian influenza is a very important taskfor practical veterinary medicine. LAMP is one of such methods.The technique is based on isothermal amplification of nucleic acids.It does not require special conditions and equipment (PCR cyclers),therefore it is cheaper in comparison with PCR. Accurate diagnosisis necessary for determining the risk associated with avian influenzain Ukraine and along the Dnipro River during the migratory season.MethodsFor the research, we used PCR-RT commercial kit Bird-Flu-PCR(Ukrzoovetprompostach, Ukraine), LAMP (the protocol has beenoptimized and patented by SSRILDVSE), QIAamp®Viral RNA MiniKit. For the study, we used pathological and biological materials frombirds, which were sent to the SSRILDVSE from all regions of Ukraineaccording to the 2013–2014 State monitoring plan.Set up of the real time PCR reactions and parameters ofamplifications are indicated in the instruction to the kit.The following protocol was used to set up the RT- LAMP: 2.5μL10 X Thermopol buffer, 1 mmol/L betaine, 5 mmol/L MgSO4,1.4 mmol/L - BNTP, 12.5μmol/L SYBR GREEN, 0.5 mmol/LMnCL2, up to 25μL Nuclease-free water, 8 U Bsm DNA polymerase,0.1μM/1 of F3, 0.1μM/1 of B3, 0.8μM/1 of FIP, 0.8μM/1 of BIP,0.4μM/1 of LF, 0.4 of LB, 2μL cDNA.During our work, we used the following optimal temperature andtime for the amplification – 59°C and 60 minutes.The sensitivity of diagnostic kit Bird-Flu-PCR and RT- LAMP wasdetermined by testing cDNA of the reference strain of AIV H5N1,which was provided to us by NSC Institute for Experimental andClinical Veterinary Medicine (Kharkiv, Ukraine). For the standard,we employed concentration in the range of 10.0-0.01 ng/sample.ResultsTable 1.This table shows the reproducibility results obtained by bothmethods. However, taken into account absence of highly pathogenicavian influenza virus circulating in Ukraine during the studied period,it was not possible to confirm these results with protocols of positivesamples.Table 2.It has been established that the sensitivity of PCR-RT kit Bird-Flu-PCR is 0.01 ng/sample for gene M and 0.1 ng/sample for subtypeH5N1.Fig. 1. Visual detection of LAMP products with differentconcentrations of cDNA of avian influenza virus (ng per sample):1 – 10; 2 – 5; 3 – 1.0; 4 – 0.1; 5–7 – 0.01; 8–9 – 0.1; 10 – negative.We have examined the LAMP results using electrophoresis forthe confirmation of visual detection and correct interpretation of theresults (Fig. 2).Fig.2. Electrophoresis results for LAMP products. M –molecular weight marker; 1 – 10.0; 2 – 5.0; 3 – 1.0; 4 – 0.1; 5–7– 0.01; 8 - negative control.It has been established that the sensitivity of LAMP is0.1 ng/sample. Slightly lower sensitivity of LAMP in comparisonto PCR-RT can be explained by visual detection of the products ofthe LAMP reaction.Conclusions1. Sensitivity of both methods is high.2. LAMP is a perspective screening method for the diagnosis ofviral infectious diseases supported by confirmation of positive resultsby PCR-RT.


2021 ◽  
Vol 2002 (1) ◽  
pp. 012075
Author(s):  
Xianchang Xi ◽  
Zhikai Huang ◽  
Lingyi Ning ◽  
Yang Zhang

2014 ◽  
Vol 12 (6) ◽  
pp. 1134-1140 ◽  
Author(s):  
Daniel Bristot de Oliveira ◽  
Romulo Silva de Oliveira

Sign in / Sign up

Export Citation Format

Share Document