Development of the method of multi-level segmentation of images
| Autor: | M. Нalushko, Y. Maistrenko, L. Tymchenko |
|---|---|
| Rok vydání: | 2019 |
| Předmět: | |
| Zdroj: | Collection of scientific works of the State University of Infrastructure and Technologies series "Transport Systems and Technologies". 34:210-216 |
| ISSN: | 2617-9059 2617-9040 |
| DOI: | 10.32703/2617-9040-2019-34-2-4 |
| Popis: | As the description of the multilevel segmentation algorithm implies, the following operations are required for its implementation. This is to determine the brightness of the pixels, their average value, and their differences. In addition, a comparison operation is required: "greater", "less", "equal" to the difference of brightness with a fixed threshold value and the operation of masking (shift) in the case of image processing by scanning it with a standard mask image. Then a summation operation is required to calculate the number of positive, negative and null drugs and the operation of calculating the maximum of their product. As is evident from the above operations, the laborious and slow computational operations inherent in, for example, orthogonal transformations are not used to implement the proposed methods. This makes it promising to apply this approach to real-time multimedia image processing, followed by their classification. |
| Databáze: | OpenAIRE |
| Externí odkaz: |
|