SnapCrack is an Android application and web UI that is capable of crack detection and localization in real time. SnapCrack leverages basic object detection algorithms to detect cracks in real-time. After a crack has been detected on the mobile application, an image will be stored to the device and sent to a database where it will be stored until it is viewed in a web portal. This repository contains this application as well as tools for modifying the dataset and training the object detection model.
1.) SnapCrack_App
The SnapCrack_App module contains the Snapcrack Android Application.
2.) Training
The Training module contains installation steps and tutorials for training a custom object detector.
3.) Server
The Server module contains basic testing for sending images to a server and database.
4.) snapcrack
The snapcrack module contains the code for the React web portal.
the SnapCrack dataset used to train the model can be found here, aswell as the corresponding tfrecord files can be found here. These can be used or re-generated to create a new model.