Research Areas

Applied Soft Computing Laboratory covers a broad variety of issues in the fields of computer vision, machine learning and their applications. Our main research interests are as following (but not limited):

  • 3D shape reconstruction (Shape from Focus)

  • Blur Detection and Segmentation

  • Image Analysis

Shape from Focus

Acquiring depth information or 3D shape of a scene is an essential task for many computer vision applications such as synthetic focus, autonomous navigation, action recognition, 3DTV and augmented or virtual reality (AR/VR). Shape from focus (SFF) is a monocular technique that employs the image focus as a cue to recover the 3D shape.

SFF can be computed in a number of steps. In the first step, a number of images for the object or scene are captured by gradually varying the focus settings of the camera. At the second step of SFF, the focus quality of each pixel in the image sequence is estimated. For this, an appropriate operator (named as focus measure (FM)) is applied on each pixel in the image sequence. The erroneous focus volume is enhanced in the third step. at the fourth step, an initial map is computed by maximizing the focus measure along z-direction. Finally, initial depth is further improved through machine learning and/or approximation methods.


Demo and Datasets

Guided Image Filtering in Shape from Focus: A Comparative Analysis, Pattern Analysis 2021 [Demo Code]


Blur Detection and Segmentation

Blur in images is considered as an undesirable effect because it leads to the loss of the necessary details required for the scene interpretation. Automatic detection of blurred and sharp pixels in an image and their classification into respective regions are very important for different image processing and computer vision applications [1]. The benefits of blur segmentation are exhibited in many applications including but not limited to object detection , scene classification, image segmentation , background blur magnification , depth of field extension and depth estimation

Blur detection and segmentation for a single image without any prior information is a challenging task. Main tasks involved in blur detection and segmentation can be depicted from the figure.