Towards a balanced trade-off between speed and accuracy in unsupervised data-driven image segmentation
| Autor: | Balázs Varga, Kristóf Karacs |
|---|---|
| Rok vydání: | 2013 |
| Předmět: |
Morphological gradient
Computer science Segmentation-based object categorization business.industry ComputingMethodologies_IMAGEPROCESSINGANDCOMPUTERVISION Scale-space segmentation Pattern recognition Image segmentation Computer Science Applications Image texture Minimum spanning tree-based segmentation Hardware and Architecture Region growing Computer vision Computer Vision and Pattern Recognition Artificial intelligence Range segmentation business Software |
| Zdroj: | Machine Vision and Applications. 24:1267-1294 |
| ISSN: | 1432-1769 0932-8092 |
| DOI: | 10.1007/s00138-013-0503-3 |
| Popis: | When it comes to image segmentation in the megapixel domain, most state-of-the-art algorithms use sampling to reduce the amount of data to be processed to reach a lower running time. Random patterns and equidistant sampling usually result in a suboptimal result because, in general, the distribution of image content is not homogeneous. The segmentation framework we propose in this paper, employs a content-adaptive technique that samples homogeneous and inhomogeneous regions sparsely and densely, respectively, thus it preserves information content in a computationally efficient way. Both the sampling procedure and the pixel-cluster assignment are guided by the same nonlinear confidence value, calculated for each image pixel with no overhead, which describes the strength of the pixel-cluster bond. Building on this confidence scheme, each pixel is associated with the most similar class with respect to its spatial position and color. We compare the performance of our framework to other segmentation algorithms on publicly available segmentation databases and using a set of 10-megapixel images, we show that it provides similar segmentation quality to a mean shift-based reference in an order of magnitude shorter time, the speedup being proportional to the amount of details in the input image. Based on our findings, we also sketch up novel design aspects to be taken into account when designing a high resolution evaluation framework. |
| Databáze: | OpenAIRE |
| Externí odkaz: |
|