Browsing by Author "Das, Debashish"

Now showing 1 - 6 of 6

Design and Implementation of a Remote Sensing Electricity Supply Control for Industrial Applications
(Faculty of Applied Science and Engineering, Shahjalal University of Science and Technology, Sylhet, Bangladesh, 2011-01-11) Bhuyan, Muhibul Haque; Das, Debashish
In this work, we have developed a remote sensing electricity supply control system for industrial applications. For this purpose, we have used a magnetic contactor, two cell phones, two voltage amplifiers, and one relay. The output voltage of the cell phone's ringer is used for sensing and providing the control signals from the remote position. Magnetic contacts and relays are used to turn on or turn off the electricity supply from the main distribution board. The circuit has been designed, implemented, and tested in various industrial areas in Bangladesh and it has been found that this controller circuit works very well.
Design and Manufacturing of a Low-Cost Synchronous Motor-Generator for Electrical Machine Laboratory
(Faculty of Applied Science and Engineering, Shahjalal University of Science and Technology, Sylhet, Bangladesh, 2011-01-11) Bhuyan, Muhibul Haque; Das, Debashish; Islam, Rafiqul
In this work, we have developed a very low-cost synchronous machine that can be run as both a synchronous motor as well an alternator. When the machine runs as an alternator, an induction motor or a DC shunt motor is required as the prime mover. This has also been developed. The machine has four poles. Its stator voltage is 415 V (AC), stator current is 0.75 A, rotor voltage is 0 to 240 V (DC) and rotor current is 0.4 A. With this machine, at least eight experiments can be performed for undergraduate-level students. There are future expansion provisions in this machine. All the experimental results support theory. It has also been found that the machine reduces the cost by around 67 % more than that of an equivalent machine imported from a foreign country.
Design of a Microcontroller Based Automatic Power Factor Improvement Plant for Power System Protection Laboratory
(Bangladesh Electronics and Informatics Society, 2014-06-26) Bhuyan, Muhibul Haque; Das, Debashish
This paper presents a microcontroller-based automatic power factor improvement plant for the power system protection laboratory of the Department of Electrical and Electronic Engineering of the Green University of Bangladesh. The load is three three-phase induction motors which cause low and lagging power factors. The microcontroller in the plant senses the phase angles of voltage and current and then decides which capacitor to turn on or which one to turn off. Accordingly, capacitors are coupled with the load from a bank of three-phase capacitors. The controller tries to set the power factor as close to unity. The capacitors are connected to or disconnected from the load through magnetic contacts and relays. Experimental observation shows that the designed power factor improvement plant works very well.
Improving Image Compression with Adjacent Attention and Refinement Block
(IEEE, 2022-02-23) Jeny, Afsana Ahsan; Islam, MD Baharul; Junayed, Masum Shah; Das, Debashish
Recently, learned image compression algorithms have shown incredible performance compared to classic hand-crafted image codecs. Despite its considerable achievements, the fundamental disadvantage is not optimized for retaining local redundancies, particularly non-repetitive patterns, which have a detrimental influence on the reconstruction quality. This paper introduces the autoencoder-style network-based efficient image compression method, which contains three novel blocks, i.e., adjacent attention block, Gaussian merge block, and decoded image refinement block, to improve the overall image compression performance. The adjacent attention block allocates the additional bits required to capture spatial correlations (both vertical and horizontal) and effectively remove worthless information. The Gaussian merge block assists the rate-distortion optimization performance, while the decoded image refinement block improves the defects in low-resolution reconstructed images. A comprehensive ablation study analyzes and evaluates the qualitative and quantitative capabilities of the proposed model. Experimental results on two publicly available datasets reveal that our method outperforms the state-of-the-art methods on the KODAK dataset (by around 4dB and 5dB) and CLIC dataset (by about 4dB and 3dB) in terms of PSNR and MS-SSIM.
Improving Image Compression with Adjacent Attention and Refinement Block
(Daffodil International University, 2022-08-01) Jeny, Afsana Ahsan; Islam, MD Baharul; Junayed, Masum Shah; Das, Debashish
Recently, learned image compression algorithms have shown incredible performance compared to classic hand-crafted image codecs. Despite its considerable achievements, the fundamental disadvantage is not optimized for retaining local redundancies, particularly non-repetitive patterns, which have a detrimental influence on the reconstruction quality. This paper introduces the autoencoder-style network-based efficient image compression method, which contains three novel blocks, i.e., adjacent attention block, Gaussian merge block, and decoded image refinement block, to improve the overall image compression performance. The adjacent attention block allocates the additional bits required to capture spatial correlations (both vertical and horizontal) and effectively remove worthless information. The Gaussian merge block assists the rate-distortion optimization performance, while the decoded image refinement block improves the defects in low-resolution reconstructed images. A comprehensive ablation study analyzes and evaluates the qualitative and quantitative capabilities of the proposed model. Experimental results on two publicly available datasets reveal that our method outperforms the state-of-the-art methods on the KODAK dataset (by around 4dB and 5dB) and CLIC dataset (by about 4dB and 3dB) in terms of PSNR and MS-SSIM.
Spectre
(Daffodil International University, 2019-12-01) Das, Debashish
In the past few years Bangladesh has experienced a tremendous growth in the field of ICT. We are becoming tech-dependent generation day by day. Though we need more development in the security sector of Smart Home to ensure more safety. As some lacking are still present in this area. There are many cases available which are related to home invasion. We can decrease the rate and avoid these misfortune situations by using advanced technology and digital devices. Traditional security systems are very weak to fight these problems as those doesn’t include advanced technology. So most of our homes are not really well secured and its not easy to stay alert about the upcoming event to avoid unwanted situation. Our goal is to make this system available for people to get the facilities of advanced security system. We will involve more ICT based application and advanced technology in our doorbell or CCTV to upgrade the security system. Our research-based project mainly focuses on predicting human intension. It will collect data by reading human facial expression and body language. Thus, our system will send alert to the user so that the user can easily detect the possibility and become aware of the situation.