| Summary: | In recent years, there were many technological advancements in the field of object detection such as face recognition and face detection. All these advancements and developments have improved work productivity and efficiency across various industries. However, in the construction industry, the use of technology has not been fully exploited to maximise work productivity and efficiency. Construction sites are still employing manual means to monitor the work progress of construction. Implementing a vision system could simplify and improve the efficiency of progress monitoring in construction sites. The vision system comprises of a detection model that incorporates the use of Convolutional Neural Network (CNN) to analyse and classify the image. This report highlights the data collection used to construct the dataset and the documentation of the training process where the detection model is trained using YOLOv3 (You Only Look Once Version 3), an object detection software. The results are evaluated and applied to calculate the work progress in construction sites. Through the implementation of a trained vision system, it could improve the efficiency in monitoring work progress and thus increasing work productivity.
|
|---|