Key research themes
1. How can edge detection be optimally designed and quantitatively evaluated to balance detection accuracy, localization precision, and noise robustness?
This research area focuses on the mathematical formulation, algorithmic design, and objective measurement of edge detectors, emphasizing the trade-offs between sensitivity to true edges, precise edge localization, and resilience to noise or spurious detections. Advancements here impact numerous computer vision tasks requiring robust edge maps for subsequent analyses.
2. How can edge-related filtering and enhancement techniques be designed to improve image restoration, sharpening, and noise reduction without introducing artifacts?
This theme explores novel algorithmic developments that leverage nonlinear filtering, edge gradient restoration, and advanced kernel designs (including dilated and extended filters) to enhance image features such as edges while suppressing noise and artifacts like ringing or halos. These techniques are crucial for improving the perceptual quality and accuracy of images in applications spanning medical diagnostics to remote sensing.
3. How do visual edge influence phenomena affect perceptual organization, filling-in processes, and aesthetic preferences in images?
This research theme investigates the perceptual and cognitive effects of edges in visual stimuli, including how edges trigger surface filling-in mechanisms, contribute to phenomena such as afterimages, modulate perceived brightness, and influence aesthetic preference for shape curvature. Understanding edge influence on perception bridges neuroscience, psychology, and computer vision.