Prominent region detection using lab color transforms and spatial support
Shaik Gousia Shabnam, Nallagarla Ramamurthy · 2017
Region detection is experiencing more and more consideration due to its rapid growth in many domains. This paper focuses on prominent region detection. Prominent region detection means detecting important parts of an image using global and local features, which complement each other to compute a prominence map. High-dimensional color transform is the method used to detect the prominent regions in an image. By mapping low-dimensional red, green, blue to a feature vector in high-dimensional color space, we can composite an accurate prominence map by finding the optimal linear combination of color coefficients in high-dimensional color space. To improve the performance of the prominent region detection a method called lab transform is proposed in this paper. From the experimental results it is clear that the performance is better compared to high-dimensional color space(HDCT).