An Automatic Brick Grading System Using Convolutional Neural Network: Bangladesh Perspective

In this work our primary objective is to devise an automated system for grading bricks based on their physical attributes. The conventional brick grading methods heavily rely on manual inspection, a process known for its time-consuming nature and subjectivity. To address this, our research proposes a machine learning-centric approach, leveraging computer vision techniques and deep learning algorithms to accurately and efficiently classify bricks. Our system is designed to extract pertinent features from brick images, focusing on properties such as color, texture, and size. The model is trained using an extensive dataset comprising various brick types. A key innovation of our work is the integration of an auto-detection system based on deep learning networks. Operating with an impressive accuracy of 89.58%, the system excels in image grading. The potential impact of our automated brick grading system extends to the construction industry, offering a time and cost-effective alternative to manual brick sorting. This abstract presents a new development in the industry and puts our work at the forefront of the development of automated brick grading systems. As far as we are aware, no one has ever completed this unique work before.