REAL-TIME OBSTACLE DETECTION USING SMART IMAGING FOR NAVIGATION OF AUTONOMOUS SMALL VESSEL
The world has moving toward unmanned machinery in many areas nowadays. The experts and scientists have gather around the world to jump into this new kind of technology development. All the technologies involve in making autonomous or unmanned machine can be very complicated as the machine need to be...
Saved in:
| Main Author: | |
|---|---|
| Format: | Final Year Project |
| Language: | English |
| Published: |
IRC
2019
|
| Online Access: | http://utpedia.utp.edu.my/20130/1/FINAL%20Dissertation%20Report_Abdul%20Hakim%20bin%20Hasane_20015.pdf http://utpedia.utp.edu.my/20130/ |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Summary: | The world has moving toward unmanned machinery in many areas nowadays. The experts and scientists have gather around the world to jump into this new kind of technology development. All the technologies involve in making autonomous or unmanned machine can be very complicated as the machine need to be working using their own artificial brain without human interruption. This artificial brain or called as deep neural network is the most important part in making an autonomous machine working successfully on their own. In industrial area, there are several tasks which had been done by human before been replaced by these autonomous or unmanned mechanical robot. The heavier or tougher task could be replaced by unmanned machine to increase production rate and human safety. In this paper, the critical part of autonomous vessel system are the vision and sensing technology. Designing the vision part will be the most crucial task as the decision-making process made by the autonomous system need to be very accurate. Smart imaging system will be focused on to detect and recognize obstacle at 8 different distance in front of the camera system. Object detection is defined as a technical process of detection on object interest and mostly done after filtering the image using several well-known methods such as canny edge detection, Sobel edge detection, Laplacian edge detection, cartoonizer technique and colour segmentation. Edge detection is very important for the object detector to be able to increase the percentage of detection of object interest. Data of images will be convert to 2 dimensional of grayscale images to enhance the edge of two different pixels intensity. |
|---|