Deepak Kumar Dewangan

ORCID: 0000-0002-0160-4215
Publications
Citations
Views
---
Saved
---
About
Contact & Profiles
Research Areas
  • Autonomous Vehicle Technology and Safety
  • Advanced Neural Network Applications
  • Vehicle License Plate Recognition
  • Traffic Prediction and Management Techniques
  • Advanced Image Processing Techniques
  • Image and Signal Denoising Methods
  • Video Surveillance and Tracking Methods
  • Anomaly Detection Techniques and Applications
  • Smart Agriculture and AI
  • Infrastructure Maintenance and Monitoring
  • Traffic control and management
  • Advanced Image Fusion Techniques
  • Brain Tumor Detection and Classification
  • Face recognition and analysis
  • Fire Detection and Safety Systems
  • Context-Aware Activity Recognition Systems
  • Industrial Vision Systems and Defect Detection
  • Automated Road and Building Extraction
  • Biometric Identification and Security
  • IoT and Edge/Fog Computing
  • Human Pose and Action Recognition
  • User Authentication and Security Systems
  • Remote Sensing and LiDAR Applications
  • Smart Parking Systems Research
  • IoT and GPS-based Vehicle Safety Systems

Atal Bihari Vajpayee Indian Institute of Information Technology and Management
2023-2024

Thapar Institute of Engineering & Technology
2024

Siksha O Anusandhan University
2022

National Institute of Technology Raipur
2011-2022

The NorthCap University
2017

ITM University
2017

Rochester Institute of Technology
2005

Artificial intelligence in vision based approaches have proven to be effective various phases of intelligent vehicle system (IVS). An IVS has intelligently take many critical decisions heterogeneous environment. Speed bump detection is one such issue real world due its varying appearance dynamic scene. The major the scaling objects from far distance and often viewed as small entity. In proposed article, deep learning computer speed model proposed, which assist control driving behavior an...

10.1109/jsen.2020.3027097 article EN IEEE Sensors Journal 2020-09-29

Abstract Advancement in vision‐based techniques has enabled the autonomous vehicle system (AVS) to understand driving scene depth. The capability of scene, and detecting specific object depends on strong feature representation such objects. However, pothole objects are difficult identify due their non‐uniform structure challenging, dynamic road environments. Existing approaches have shown limited performance for precise detection potholes. study potholes, intelligent behaviour is little...

10.1049/ell2.12062 article EN cc-by Electronics Letters 2020-12-31

Lane detection on the road is most essential phase of existing advanced driver assistance systems (ADAS). In current scheme, lanes are parallel tracks which marked based different speed limit for vehicles. markings defined as driving regions vehicles in order to avoid a collision. an Intelligent Vehicle System (IVS), these being identified by various massive computing techniques and sensor-based mechanisms like radar, LiDAR, GPS have high operational costs. However, vision-based methods...

10.1109/jsen.2020.3037340 article EN IEEE Sensors Journal 2020-11-11

The recent advancement in artificial intelligence approach or deep learning techniques explored the ways to facilitate automation various sectors. application of with computer vision field has resulted realization intelligent systems. Vehicle detection plays a key role Intelligent System and Transport as it assists critical components these systems like road scene classification, detecting obstacle vehicles find an unhindered pathway, even preventing accidents. This paper presents...

10.1109/iciccs51141.2021.9432374 article EN 2022 6th International Conference on Intelligent Computing and Control Systems (ICICCS) 2021-05-06

“Road detection is said to be a major research area in remote sensing analysis and it usually complex due the data complexities as gets varied appearance with minor inter-class huge intra-class variations that often cause errors gaps extraction of road”. Moreover, majority supervised learning techniques endure from high price manual annotation or inadequate training data. Thereby, this paper intends introduce new model for road detection. This work exploits siamesed fully convolutional...

10.1142/s0218001422520024 article EN International Journal of Pattern Recognition and Artificial Intelligence 2022-04-22

Detection of speed bumps is a challenging issue in autonomous ve-hicle systems due to their appearance dynamic environments. The unmarked bump not an apparent entity even the daytime, thus carries difficulty precisely detect for vision-based driving assistance system (DAS). In some cases, DAS may misinterpret marked and zebra crossing pattern, which can further weaken recognition performance risk human safety. this letter, convolutional neural network model using region proposal nonmaximum...

10.1109/lsens.2024.3360095 article EN IEEE Sensors Letters 2024-01-30

One of the most difficult challenges in field computer vision is human activity recognition (HAR). The main purpose intelligent video system to determine actions and activities individual. This action monitoring can be used a variety settings, which include human-computer interaction, tracking, security, health monitoring. Detecting an uncircumscribed territory remains challenging task with multiple challenges, despite ongoing efforts this area. Throughout article, some recent research...

10.1109/iccmc51019.2021.9418255 article EN 2022 6th International Conference on Computing Methodologies and Communication (ICCMC) 2021-04-08

Traffic obstruction is one of the major problems faced by most metropolises in dynamic road scene. Most roads various cities (especially India) are poorly designed and lack separate footpaths many reasons for traffic congestion. lights deployed at intersections inefficient, causing issues including waiting time vehicles energy wastage. In proposed approach, a learning model using deep reinforcement technique to solve issue managing light cycle has been focused. The collect data from through...

10.1109/icais50930.2021.9395880 article EN 2021-03-25

Contrast improvement are being aided by the universal data content of an input image enlarging dynamic variety intensity levels, utilized Conversion functions. Certain conversion functions utilize local substantial for modifying details, such as quality & boundaries. In this paper, effective method improving contrast, is introduced, in which a plotting utility, blend and functions, preserves fine facts addition. widening Image preserved completely function. The instinctive procedure...

10.1109/iccsp.2017.8286525 article EN 2017-04-01

The density of current road traffic frequently producing circumstances that endanger the protection individuals. For safety individual, one has to keep safe distance from an upcoming vehicle in a busy road. In intelligent system, selection sensors, orientation and velocity persons other moving objects are significant for logical decisions given scenario. These acquired information can be deployed offer Driver Assistance System (DAS) future vehicle. proposed work, vision-based sensors castoff...

10.1109/icpc2t48082.2020.9071478 article EN 2020-01-01

Technology development has altered how people live. Intelligent vehicles are being developed as a result of technological advancements in the automobile industry. Pedestrians most vulnerable road users, accounting for more than half all traffic fatalities. Thus, key thing to take into account is pedestrian detection. In order safeguard both drivers and pedestrians, an intelligent vehicle system can detect pedestrians front moving either loudly warn driver risk or automatically slow down...

10.1109/otcon56053.2023.10113992 article EN 2023-02-08

Movement Planning, as a principal innovation of programmed routes for Autonomous vehicles, is yet an open testing issue, all things considered, traffic circumstances and generally applied by the model-based methodologies. Nonetheless, because multifaceted nature vulnerability edge cases, it difficult to devise overall movement arranging framework vehicles. Spurred this expanded fame, we give Deep-learning based ways deal with vehicle motion prediction practically 80% Accuracy in paper. The...

10.1109/iccmc51019.2021.9418449 article EN 2022 6th International Conference on Computing Methodologies and Communication (ICCMC) 2021-04-08

Satellite Images are used nowadays in multiple applications. Generally these very noisy and blurry. This paper presents comparison of different techniques for contrast enrichment on the basis their PSNR values. In method proposed by us, removal noise blur instinctive procedure suggested pre-processing it also lowers consequence noise. Contrast improvement is being aided universal data content an image enlarging dynamic variety intensity levels, utilized Conversion functions. Fuzzy based...

10.1109/icecds.2017.8389506 article EN 2017-08-01

Autonomous parking capability is one of the major applications autonomous vehicles in automobile industry. The standard vehicle requires acceleration, braking, and steering control to park into particular space. However, a small amount driving error any mentioned towards may produce hazardous situations lead road accidents. Apart from scenario, parallel reverse are too risky application areas which controlling computation has be very specific. For demanding market vehicles, this pin-point...

10.1109/iciccs51141.2021.9432362 article EN 2022 6th International Conference on Intelligent Computing and Control Systems (ICICCS) 2021-05-06

This Satellite Images are used nowadays in multiple applications like astronomy, weather forecasting, Space Sciences and earth other planet observation. Generally very noisy blurry. In this paper we presenting different techniques for satellite image enhancement. For removal of noise blur instinctive procedure suggested that pre-processes poorly concentrated or degraded images by composing numerous consecutive autonomous processing phases which overturn noise, enrich contrast & upturn the...

10.1109/icecds.2017.8390024 article EN 2017-08-01
Coming Soon ...