Development of pipes counting application for warehouse inventory based on Android platform

Zairul Akhbar, Nor Fatin Syahirah (2024) Development of pipes counting application for warehouse inventory based on Android platform. Project Report. Universiti Teknikal Malaysia Melaka, Melaka, Malaysia. (Submitted)

Text (Full text) Development of pipes counting application for warehouse inventory based on Android platform.pdf - Submitted Version Download (1MB)

Abstract

The project aims at the development of a pipes counting application designed for warehouse inventory based on the Android platform. This study focuses on pipes, which are crucial elements in construction materials at construction sites and employs deep learning technology to automate the pipe counting process. The design of a mobile application is a part of the development of pipe counting applications which enables warehouse workers to effectively count and manage the stock of pipes in its inventory. Inaccuracy of inventory records appears to be a widespread issue, particularly in industrial and residential sectors. While manual counting is a widespread practice at the warehouse, it is impractical, challenging, and time-consuming. The process of counting the pipes manually often takes several hours for workers. Thus, human error can take place due to the manual counting process. To address the challenge, a pre-trained YOLOv5 model based on the PyTorch framework is used to prepare the data model training. The REST API server with the framework of Flask to expose the data model training has been integrated into the Flutter framework deployed on mobile applications for testing. The application prompted users to input the image of a captured pipe. The response will be processed from the YOLOv5 REST API in the Android app to return the detected objects and their coordinates. The detected objects and their counts will be displayed on the Android app’s user interface.

Actions (login required)

View Item