Our Projects
Minor Project : “DIGITAL ATTENDANCE SYSTEM”
Digital Attendance System is a system, which takes attendance of attendees using face recognition techniques using video input. The record of an individual is very important to organizations such as colleges and institutes. The conventional methodology for taking attendance is by calling the name or roll number of the student and then recording the attendance. This type of traditional system of keeping an individual attendees record so is more tedious and time consuming. To overcome this problem we developed “Digital Attendance System” which keeps the record of an individual student by recognizing their faces. It can be sayed that the uniqueness or individuality of an individual is his face and hence the main goal of this project is to recognize faces in real time and environment for educational institutes or organization to mark the attendance of individual attendees on daily basis and keep track of their presence. This system is automated which takes photos from video input which are then compared with photos that are initially stored in database and accordingly mark attendance and finally generating attendance report. The system uses Haar cascades for detecting haar-like features to detect faces, PCA from OpenCV to recognize faces and SQLite for Database operations for attendance marking purpose.
Project Members
Submitted To
DEPARTMENT OF COMPUTER AND ELECTRONICS & COMMUNICATION ENGINEERING
LALITPUR, NEPAL
Major Project : “Short Term Flood Prediction Using Machine Learning”
Floods are one of the most destructive natural disasters which cause damage to lives, properties and possessions as well as disruption to communication and other important factors. As of this time there is no mechanism to avoid flood but its prediction can secure the live of inhabitants and also can reduce damage. This project presents the short term flood prediction model which will predict the river floods using machine learning approach. This model uses water level, rainfall, soil moisture and some other factors for the prediction. Though high number of factors may affect flood in rivers only some of the most responsible factors are considered in this model. The main aim of the project is to calculate probability of flood in the river for up to one week in advance using a hybrid ANN model between back propagation and OWO-HWO (Outer Weight Optimization-Hidden Weight Optimization) which takes information from hydrological stations as inputs. Though the prediction is very hard to model for floods, minimal error and high reliability in the result are expected. To validate the results some error indicators such as MAE, MSE, RMSE, EF, CD can be calculated.
Project Members
Submitted To
DEPARTMENT OF COMPUTER AND ELECTRONICS & COMMUNICATION ENGINEERING
LALITPUR, NEPAL