چكيده به لاتين
The railway level crossing is one of the most important and high-risk points of the railway. So it is very important to improving safety in these areas. Because of converting all level crossings to non-level crossings requires a lot of time and money, it is better to use mechanized equipment called intelligent barrier system to safely control these intersections to eliminate the risks of human error from the system. To be. In such a system, the existence of an obstacle detection tool is required.
In this project, "image processing techniques" have been used as an intelligent tool. For this purpose, in Python programming environment and by implementing image processing techniques, the obstacles were well revealed so that the obstacles characteristics such as their material, size and color do not cause system malfunction. Then, an attempt was made to eliminate errors in light conditions (night and day) and different weather conditions (snow, rain, fog, etc.) and a program was developed with the ability to detect obstacles in different conditions. In order to increase the processing speed, the danger zone was defined, so that only the data in this area are processed and other non-useful data such as the movement of barriers and foliage of trees are not considered. Therefore, the volume of data decreases and the detection speed increases. In the second part of the program, by using neural networks, the type of obstacle (human, animal, vehicle, etc.) is specified. This leads to the division of obstacles into two groups of dangerous obstacles and safe obstacles.
