This thesis is focused on developing a robust algorithm for recognition of road signs including all stages of a machine vision system i.e. image acquisition, pre-processing, colour segmentation, labelling and classifi-cation. Images are acquired by two different imaging systems and noise removal is done by applying Mean filter. Furthermore, different colour segmentation methods are investigated to find out the most high-performance approach and after applying dynamic segmentation based on blue channel in YCbCr colour space, the obtained binary image is transferred to a personal computer through the developed PC software using standard serial port and further processing and classification is run on the PC. Histogram of Oriented Gradients (HOG) is used as the main feature for recognition of road signs and finally the classification task is fulfilled by employing hardware efficient Minimum Distance Classifier (MDC).
| Identifer | oai:union.ndltd.org:UPSALLA1/oai:DiVA.org:miun-17370 |
| Date | January 2012 |
| Creators | Hashemi, Ashkan |
| Publisher | Mittuniversitetet, Institutionen för informationsteknologi och medier |
| Source Sets | DiVA Archive at Upsalla University |
| Language | English |
| Detected Language | English |
| Type | Student thesis, info:eu-repo/semantics/bachelorThesis, text |
| Format | application/pdf |
| Rights | info:eu-repo/semantics/openAccess |
Page generated in 0.0019 seconds