Edge detection; Pixel classification; Region segmentation; Thresholding
The rapid rate of image analysis field has grown enormously in the past few decades. Image analysis intends to construct explicit, meaningful descriptions of physical objects in images. It can be divided into two parts: low-level image analysis and high-level image analysis. Low-level tasks focus on region-based segmentation, whereas high-level tasks are related to object-oriented representation. Image segmentation, a process of pixel classification, aims to extract or segment objects or regions from the background. Intrinsic images can be generated at the low-level processing, revealing physical properties of the imaged scene. This can often be implemented with parallel computation.