AMC Bridge using ML to spot hard-hat safety breaches

Hard-hat Detector technology demonstration designed to improve safety on construction sites

AMC Bridge is working on a prototype web-based technology that applies machine learning to identify workers that do not wear hard hats on a construction site.

Construction Safety—Hard-hat Detector uses machine learning and image recognition techniques of the TensorFlow open-source software library to analyse frames extracted from an uploaded video.

Once the video file is processed, it is possible to view snapshots where non-hard hat wearers are detected or download them locally. The technology demonstration can be ‘seamlessly connected’ to Google Drive, Dropbox, or MEGA online.