Font size

  • S
  • M
  • L

Laboratory of Media Dynamics
Graduate School of Information Science and Technology
Hokkaido University

Image Restoration

Missing Area Restoration Technology

Missing area restoration is applied in various fields. For example, removal of unnecessary objects, restoration of corrupted old films and error concealment of videos transmitted in error prone environment are representative applications. Such methods enable the restoration of images to the point where the human eye no longer suspects that the images were ever degraded in the first place.

Deblurring Technology

Whenever we capture an image, we often unknowingly cause various degradations, such as blurring due to poor focusing or camera shake. The conventional solution to this problem is to equip the camera with functionality that prevents such degradations. More recently, it has become possible to utilize image processing technology in order to restore blurry images after they have been captured.

Impulse Noise Reduction Technology

Images contain various types of noise. Impulse noise has a particularly strong negative effect on image readability. Our technology allows the removal of impulse noise without excessive smoothing of edges in the image. Our technology can estimate the original, high-quality image from a degraded image that is nearly unreadable.

Fog Removal Technology

Reduced visibility during bad weather conditions severely impacts road safety. We have proposed a method of applying image processing to reduce the effects of strong fog and improve visibility during bad weather conditions. In comparison to conventional methods, our method achieves a greater improvement in visibility.

Color Compensation Technology

Image data is usually represented using three components: red (R), green (G) and blue (B). However, real-world objects contain much more color information (optical spectrum) than that. In our research, we take an ordinary image captured by a digital camera and estimate its real-world optical spectrum. This enables various applications, such as the recovery of the true color composition from visible light images; detection of plants; and applications related to color transformations due to physical phenomena.