Deep learning combined with homography based vision algorithm for trench dimension estimation | |
| Author | Aung Zar Lin |
| Call Number | AIT RSPR no.DSAI-23-02 |
| Subject(s) | Deep learning (Machine learning) Trenchs--Safety measures--Data processing |
| Note | A research study submitted in partial fulfillment of the requirements for the degree of Master of Science in Data Science and Artificial Intelligence |
| Publisher | Asian Institute of Technology |
| Abstract | One of the crucial components of the waterworks authority is the installation of water pipes. Water pipes that require less maintenance and last longer should be installed prop erly and in accordance with industry standards. To install a water pipe, the construction procedure begins by digging a trench that has a specific width and depth. Trench-digging errors, such as using the wrong width and depth for the trench or using incorrect trench profiles, will result in water pipe damage. That might eventually result in water pipe leaks, which are very challenging to find and correct. In my research, a single camera will be used to capture an image of the trench. The YOLO v4 object detection algorithm is used to detect the trench profile. A homography matrix calculation is performed to convert image coordinates to real world coordinates. There are many different ways to calculate a homography matrix. In this research, I create the homography from co-planar points. Using the homography, the dimensions of the trench can be estimated using Euclidean distance between points P1 and P2. The algorithm received the error less than 10 percent when the camera distance is less than 800 cm and the camera angle is less than 20 degrees. |
| Year | 2023 |
| Type | Research Study Project Report (RSPR) |
| School | School of Engineering and Technology |
| Department | Department of Information and Communications Technologies (DICT) |
| Academic Program/FoS | Data Science and Artificial Intelligence (DSAI) |
| Chairperson(s) | Dailey, Matthew N. |
| Examination Committee(s) | Mongkol Ekpanyapong;Chaklam Silpasuwanchai |
| Scholarship Donor(s) | AIT Fellowship |
| Degree | Research Studies Project Report (M. Sc.) - Asian Institute of Technology, 2023 |