Project Report

Browse

Search Results

Now showing 1 - 10 of 26

Watermelon leaf disease detection and classification using Yolo architecture
(Daffodil International University, 2025-01-14) Hasan, Mir Saem
Watermelon is a fruit that people all over the world enjoy. However, eating watermelon is mostly in the summertime because the juicy and cooling effect to the body it has is highly appreciated. In countries like Bangladesh and other countries with long, hot summers, the demand for watermelon fruit is more or less similar. The yield quality is compromised due to several leaf diseases affecting watermelon production and causing a decrease in value of financial returns. Thus, very soon and accurate diagnosis of these diseases becomes very critical in order to minimize the losses and ensure sustainable agriculture. The deep learning-based approach of the study uses the YOLOv11 model for real-time detection and classification of watermelon leaf diseases. For that, a balanced dataset of healthy and diseased watermelon leaf images was collected and further added to for improvement of model performance. The specifications are that its detection speed and accuracy is very good while being lightweight in its design; because of that, the YOLOv8 architecture was selected. The model's robust precision under challenging field conditions in identifying multiple classes of diseases was attained with optimal learning parameters. Precision was achieved to be 93.3%; Recall 87.4%, mAP50, and mAP50-95 were 96.4% and 81.5% respectively. A web- based application was developed for realtime disease detection in uploaded leaf images to ensure easy reach and use for farmers and agricultural experts. The system presented in this integrated approach contributes promise toward smart agriculture through enhanced crop monitoring in favor of food security and economic stability.
Face ATT: A face recognition attendance system
(Daffodil International University, 2025-01-18) Vikki, Shadab Pervaz
With the advancement of modern technologies areas related to robotics and computer vision, real time image processing has become a major technology under consideration. So here a try has been made for a novel approach for capturing images from the Web Camera in real time environment and process them as we are required. This project portrays a machine learning approach for face recognition to accomplish this process very quick with high identification rates using OpenCV. Here in this project depicts a basic and simple equipment execution of face location framework utilizing Web Camera. The framework is modified utilizing Python programming language. The destinations of the face recognition are to recognize appearances and its spatial area in any pictures or recordings. The proposed framework distinguishes the faces present in a grey scale and colored image. This project center around usage of face detection framework for human recognizable proof in light of OpenCV library with python. Here in this project the idea of identification has been built up by composing distinguishable code for dataset generator, trainer and indicator. Effectiveness of the framework is examined by ascertaining the Face recognition rate for every one of the databases. The outcomes uncover that the proposed framework can be utilized for face detection even from low quality pictures and shows incredible execution level. At last, the data that will be shown alongside recognized photograph has been put away on database. This concept has a higher scope on security and surveillance projects and various operation. The main purpose of this project is to build a face recognition-based attendance monitoring system for educational institution to enhance and upgrade the current attendance system into more efficient and effective as compared to before. The current old system has a lot of ambiguity that caused inaccurate and inefficient of attendance taking. Many problems arise when the authority is unable to enforce the regulation that exist in the old system. The technology working behind will be the face recognition system. The human face is one of the natural traits that can uniquely identify an individual. Therefore, it is used to trace identity as the possibilities for a face to deviate or being duplicated is low. In this project, face databases will be created to pump data into the recognizer algorithm. Then, during the attendance taking session, faces will be compared against the database to seek for identity. When an individual is identified, its attendance will be taken down automatically saving necessary information into a excel sheet. At the end of the day, the excel sheet containing attendance information regarding all individuals are mailed to the respective faculty.
Automatic Number Plate Detection with Image Recognition System
(Daffodil International University, 2025-11-30) Ahamed, Sabbir
ANPR systems are very important fo traffic system safe and under control.To make these systems work well, you need to make models that are both quick and correct.This study examines the application of YOLOv8, a contemporary object detection model for the analysis of license plates in Bangladesh.We took pictures of cars in a lot of different places and styles.The dataset taught five different versions of the YOLOv8 model: YOLOv8n, YOLOv8s, YOLOv8m, YOLOv8l, and YOLOv8x.We are used standard metrics like recall, F1 score, and mean Average Precision (mAP) to find out how well the models worked.The result are showed that YOLOv8 could read license plates from Bangladesh correctly.The results show that YOLOv8 can deal with the special things about Bangladeshi plates.The study also helps you pick a model based on how fast and powerful you need your computer to be.We will add these models to a complete ANPR system and improve the dataset in the future to make traffic management and safety better
Bridging the Communication Gap: An AI-Driven word-level Bangla Sign Language Interpreter
(Daffodil International University, 2025-09-16) Rahaman, Md Saidur
The present paper is a research project report creating a real-time Bangla Sign Language (BdSL) recognition system in a two-phase approach. The initial phase concentrated on developing a state-of-the-art baseline with our proposed hybrid Conv1D-BiLSTM architecture including the Attention mechanism with an accuracy of 97.3% on the published 60-word BDSLw60 v2 dataset, which is higher than the previously published benchmarks. The second stage of research was devoted to the important issue of vocabulary development, and a new set of signs of 39 signs, which had been collected by hand, was presented. In order to address the dramatic accuracy drop experienced with first fine-tuning, we designed and adopted a professional quality multi-stage fine-tuning policy that includes class weights and gradual unfreezing. This sophisticated procedure was very practical and the final and strong model reached 93 percent total accuracy in the whole vocabulary (99-word). This last model is implemented on a real-time Flask web application, which translates signs directly to text, showing an end-to-end cycle between a validation of the baseline to a scalable and real-world application.
Emotion That Speaks: Peering Beyond the Obvious with Deep Learning for Emotion Recognition
(Daffodil International University, 2024-07-13) l Hossen, Md Shaki; Shimul, Nazmul Islam
Human emotions are spontaneous mental states produced by changes in facial muscles, leading to expressions. In various human-computer interaction applications, techniques for nonverbal communication like facial expressions, eye movements, and gestures are employed. Facial emotion, in particular, is widely utilized for conveying an individual's emotional states and feelings. However, emotion recognition is challenging due to the need for a clear distinction between facial expressions and the complexity and variability of emotions. Conventional machine learning algorithms frequently have difficulties in accurately recognizing emotions since they heavily depend on humangenerated elements. To address this issue, we explored the use of deep learning models for emotion detection based on facial expressions. Specifically, we evaluated Vision Transformer (ViT), VGG19, InceptionV3, EfficientNet, and ResNet50 models. The findings of our study demonstrated that Vision Transformer (ViT) achieved the highest accuracy rate of 82.96%, followed by Efficient-Net at 82.36%, ResNet50 at 80.87%, InceptionV3 at 79%, and VGG19 at 78.22%. Based on its excellent accuracy and robustness, we propose using the Vision Transformer (ViT) for the identification of six distinct emotions: anger, neutrality, happiness, sadness, disgust, and surprise.
ARDUINO Based Pick and Place Bot Using Image Detection
(Daffodil International University, 2024-07-13) Bappy, Md. Aslamuzzaman
The contemporary landscape of robotics is marked by the integration of smart technologies to enhance autonomy and utility. This thesis presents the development of an Arduino-Based Pick and Place Bot employing image detection, with mobile app control for versatile applications. The robot is equipped with a mobile-controlled mechanism, allowing users to navigate the bot through a mobile app interface. The pivotal feature lies in the real-time monitoring facilitated by a camera mounted on the robot, providing live feedback of its surroundings to the mobile app. The hardware foundation of this project comprises an Arduino board, motors for movement, and a camera for live imaging. The software infrastructure involves the Arduino Integrated Development Environment (IDE) and a mobile app development platform. The image detection algorithm implemented on the robot enables it to recognize and respond to objects in its vicinity. Through the mobile app, users can direct the robot's movements and collect objects from the environment. The realtime monitoring feature ensures a seamless and intuitive interaction, allowing users to make informed decisions based on the live visual feed. This thesis delves into the methodology of constructing the pick and place bot, detailing the hardware and software components, and the integration of image detection. The implementation phase includes a step-by-step guide, code snippets, and highlights of challenges encountered and solutions devised. Results demonstrate the robot's successful navigation, object recognition, and the efficacy of mobile control. In conclusion, this project contributes to the evolving field of robotics by presenting a practical application of Arduino-based automation, showcasing the potential for mobile-controlled pick and place bots with image detection capabilities in various contexts.
Sentiment analysis from fascial image:
(Daffodil International University, 2024-07-24) Shorna, Habiba Suraya
Now-a-days Face recognition system has become faster form of biometric system. It is a object detection system which is capable of recognizing face technically from digital image. Here in this paper, I used the latest version of YOLO algorithm that is YOLOv8 which is very capable of detecting real time object detection to detect multiple emotion of human faces in a single digital image. The purpose of this study is to find out how YOLOv8 can perform real time face detection accurately. Here I used four type of emotion of human face : happy, sad, angry and confused. For this study, dataset made from raw images taken from various people and annotated them and made dataset. Total of 209 images are used to make the dataset to achieve the accuracy. This study shows how accurately YOLOv8 works to perform real time face detection
A Computer Vision Approach for Banana Leaf Disease Classification
(Daffodil International University, 2024-07-13) Akash, Anik Dash; Hossain, Md. Iftekhar
This study investigates the potential of computer vision approach for classifying banana leaf diseases - Cordana, Pestalotiopsis, Sigatoka in the context of regional variations. We collected a dataset of banana leaf images from Habiganj, Sylhet, Bangladesh, to account for these variations and employed data augmentation techniques to enrich its size and diversity. Five pre-trained CNN architectures (VGG19, VGG16, ResNet50, MobileNetV1, MobileNetV2) were evaluated for their disease classification performance. The evaluation compared the models' performance with and without image preprocessing. Our findings highlight the exceptional performance of MobileNetV2, achieving an impressive 94.41% accuracy on raw, unprocessed data. This accuracy further improved to 96.41% after image preprocessing, demonstrating the model's robustness and resilience to variations that might be encountered in real-world applications. These results emphasize the potential of CNNs, particularly lightweight architectures like MobileNetV2, for accurate banana leaf disease classification using computer vision. This study provides valuable insights for developing future plant disease identification systems, ultimately contributing to improved disease management and promoting sustainable agricultural practices.
Traffic Sign Detection and Recognition Using Deep Learning
(Daffodil International University, 2024-07-13) Islam, Md. Jaharul
Automatic traffic sign detection and recognition is a crucial part of traffic management. It offers a precise and efficient method for managing traffic sign with the least amount of human effort. Incorrect direction is causing an increasing number of accidents in today's world. These accidents can be decreased with automated traffic sign detection. A vast majority of existing approaches perform well on traffic signs needed for advanced driverassistance and autonomous systems but they don't work with Bangladesh and they don't work with different environments . In our country there are hundreds of traffic signs. I work on 15 signs that are very essential for advanced driving systems and the vehicle driver. For building our model to detect object traffic signs I use YOLOv8 (You Only Look Once), VGG16 (Visual Geometry Group), Inception v3, With YOLO's assistance, machines can quickly recognize objects and forecast the traffic sign automatically. In this research best performing model is YOLOv8 and the accuracy is 96%. The main objective of this report is to show drivers how to drive safely by using the machine's instructions to anticipate all traffic regulations and directives.
Classification of Bangladeshi Bird Species Based on Images Using Deep Learning Techniques
(Daffodil International University, 2024-07-24) Sourov, Md. Shahadat Hossain
Bird species classification is significant for the ecological studies and the biodiversity conservation. This paper introduces a novel method using deep learning to automatically classify the bird species from the images. We study images of birds and apply the convolutional neural network to examine the basic features like feather and beak shape. This paper deals with the bird classification using deep learning methods by using a novel dataset coming from Kaggle containing the images of the eighteen native bird species in Bangladesh. The dataset, which is made up of 2704 carefully preprocessed images, goes through the rigorous pre-processing steps and is then divided into different subsets for the purpose of model evaluation, which includes training, testing and validation sets. Different deep learning architectures, such as MobileNetV2, DenseNet201, VGG19, and Xception, are carefully studied individually to check their effectiveness in the task of bird classification. Besides, hybrid models, which combine the different architectural paradigms, are built in a way to find the synergies and thus the classification accuracy is improved. A large number of evaluation metrics such as accuracy, precision, recall and F1-score are used to precisely compare the performance of the models. Findings reveal nuanced intricacies in model performance: The MobileNetV2 gets an accuracy of 97. 15%, VGG19 87.22%, DenseNet201 98.27%, Xception 98.33%, and the Hybrid MobileNetV2 and DenseNet201 model appears as the winner, with the highest accuracy of 99.50%. This superb achievement shows the possibility of hybrid architectures to beat the single-model approaches, giving us information how the interrelation between architectural complexity, computational resources, and task-specific requirements

Project Report

Browse

Filters

Settings

Sort By

Results per page

Search Results