Browsing by Author "Islam, Merajul"

Now showing 1 - 3 of 3

BongoAgro: a mobile solution for postharvest loss reduction fruits and vegetables
(IUB, 2025-09) Roknuzzaman, Md.; Islam, Merajul
The fruits and vegetables sector of Bangladesh suffers postharvest losses estimated at 25-40% of production, the economic and nutrition consequences of which are huge. Such losses are promoted by inefficient cold storage, poor transportation, handling, and access to market information. To overcome these challenges, BongoAgro was developed in the form of a mobile application with mapped digital tools that small holder farmers could have integrated use. With the help of the app, the real-time update is available on the market price, weather forecast, cold chain management support, and expert advisory services are within reach of fishers in a Bengali language-friendly interface. This research was conducted as a mixed-methods study, utilizing a literature review, stakeholder interviews, and a potential pilot field trial in select districts to evaluate the implementation/ease of use, and utility of the application. The findings have shown that an informed decision and efficient postharvest management could help BongoAgro mitigate the spoilage by 20% and increase the farmer’s income by 10-15%. This paper shows how mobile-based interventions can be used to enhance the agricultural supply chain and improve the lifestyle of people living in developing economies.
Daffodil Educational Smart City 2040
(Daffodil International University, 2018-11) Islam, Merajul; Barua, Tuhin; Hasan, Istiak; Islam, Md. Shariful; Raihan, Muhammad Asif; Bhuiyan, Md. Motiur Rahman
3D visualization is the process of generating three-dimensional moving images in a digital environment. Careful creating of 3D models or objects is carried out within 3D software for exporting picture sequences giving them the illusion of animation. Animation is important for all above the sector. Like: Education sector, architecture sector, historical sector etc. That’s why we choose visualization of 3D animation and also view detailing every models with “Augmented Reality” and “Virtual Reality”. From the beginning of our university life, we have a group. which group attached with a professional workflow. In our final project, we decide to do something for our university that's why we choose to do 3d visualization of Daffodil International University, permanent campus(Ashulia) by following the master plan of Daffodil International University 2040. The work that will have all our project implemented. We start DIU Educational Smart City from the masterplan AutoCAD layout, complete the model using by Max, Maya, and SketchUp. Lumion plays a role competing render the whole project and unity takes for Virtual Reality and Augmented Reality. And the building track by Adobe After Effect and finally, composition of the animation was done using Premiere Pro. Music and voice over were adopted for the presentation of the project in a professional manner. In this project, we can find an overview architectural visualize of Daffodil Educational Smart City 2040 with 3D Animation and also view Augmented Reality(AR) and Virtual Reality(VR).
Fake Hilsa Detection Using Machine Learning Algorithm
(Daffodil International University, 2021-06-01) Islam, Merajul; Rahman, Abdur; Ani, Jannatul Ferdous
Hilsa is the national fish of Bangladesh. Bangladesh is earning a lot of foreign currency by exporting this fish. Unfortunately, in recent days some unscrupulous businessmen are selling fake Hilsa fishes to gain profit. The Sardines and Sardinella are the most sold in the market as Hilsa. The government agency of Bangladesh, namely the Bangladesh Food Safety Authority said that these fake Hilsa fish contain high levels of cadmium and lead which are detrimental for humans. In this research, We have proposed a method that can readily identify original Hilsa fish and fake Hilsa fish. Based on the research available on online literature, we are the first to do research on identifying original Hilsa fish. We have collected more than 16,000 images of original and counterfeit Hilsa fish. To classify these images, we have used several deep learning based models. Then the performance has been compared between them. Among those models, DenseNet201 achieved the highest accuracy of 97.02%.