Vegetation Detection for Traffic Applications
- Marko Subasic
- Sven Loncaric
- Faculty of Electrical Engineering and Computing, University of Zagreb, Croatia
The project lasted in the second half of 2011 and was cofinanced by The Business Innovation Center of Croatia (BICRO).
Regular maintenance of railroads consists of weed elimination among other activities. Every year maintainers of railroads have to make sure that no significant vegetation develop in the predefine radius around the rails. the reason is that dense enough vegetation could be ignited from the sparks produced by passing trains which could start an open fire. Also vegetation on rails could interfere with breaking of the train. The most common way of weed elimination is spraying of herbicides along the tracks from the specialized train composition. There are different modes of operation but the one of interest concerns spraying herbicides just on areas along the tracks that are covered with vegetation. Terning sprayer on and off can be done manually but better efficiency and precision can be obtained with automatic control of spraying nozzles. The automatic control must be guided by automatic visual inspection of the area around the tracks. Precise application of herbicide enables savings in herbicide, as well as reduction of unnecessary environment pollution. Additional benefits can be obtained if different types of vegetation could be discriminated, as different types require different strengths of herbicide.
The goal of the research project was to develop an image analysis method for detection and recognition of vegetation alongside railroad tracks. Images for analysis are obtained from a video camera monitoring the immediate surrounding of the railroad tracks, mounted on the train. The detection of vegetation is based on the color information while recognition of different types of vegetation is based on texture features. The goal of detection is to identify image regions covered with vegetation. The recognition has to discriminate narrow-leaved and wide-leaved vegetation in the detected areas. The method has been designed to operate in real time in good lighting conditions.