Applied integral intensity projection to find the numbers of the parking spots
| Autor: | A. Moungkhaodaeng, P. Choorat, C. Klimanee, A. Charoensuk, V. Kaosaiyananda, B. Thepsathit |
|---|---|
| Rok vydání: | 2017 |
| Předmět: |
050210 logistics & transportation
business.industry 05 social sciences ComputingMethodologies_IMAGEPROCESSINGANDCOMPUTERVISION 02 engineering and technology Image segmentation Image (mathematics) Region of interest 0502 economics and business Vertical direction 0202 electrical engineering electronic engineering information engineering Parking lot 020201 artificial intelligence & image processing Computer vision Integral projection Artificial intelligence Projection (set theory) business Intensity (heat transfer) Mathematics |
| Zdroj: | KST |
| DOI: | 10.1109/kst.2017.7886114 |
| Popis: | The automatic parking system using an integral intensity projection is important in order to reduce the time spend for finding an available parking space. We applied the intensity projection and peak detection approaches to find the number and location of the parking space from a parking lot image. The parking lot image is firstly applied the integral projection in order to sum and smooth intensity projection of this image in a horizontal direction. After that the image is segmented into the region of interest (ROI) image by using the location object's edges derived from a derivative method. Then the integral intensity projection is again utilized the maximum points in vertical direction of the image intensity summation. Finally, the number of parking spots were counted by using peak detection approach. In the experimental results, the accuracy for counting the number of cars in the parking lot image is 53.03%. |
| Databáze: | OpenAIRE |
| Externí odkaz: |
|