Loading...

Frame Rate Upconversion Using Image Registration Techniques on Echocardiography

Haghparast, Sobhan | 2019

623 Viewed
  1. Type of Document: M.Sc. Thesis
  2. Language: Farsi
  3. Document No: 51618 (05)
  4. University: Sharif University of Technology
  5. Department: Electrical Engineering
  6. Advisor(s): Fatemizadeh, Emadoddin
  7. Abstract:
  8. In this research a novel method is proposed which uses the concept of image registration to forecast frames as a new frame up conversion method. This frame rate up conversion method is specially designed for medical videos because of the medical structure of the none rigid image registration used as a field to forecast new frames which causes better results for these kinds of videos. In this method we interpolate new frames in the field of none rigid transformation grid in order to improve video parameters such as blocking, blurriness, etc. which can help doctors to detect abnormalities in such videos much preciously. By analyzing existing methods, we can find that this field is bereft of a method which uses medical imaging techniques to get better result. as we know the mediums in medical images are none rigid and it can’t be a good idea to analyze these images without considering their characteristics.in some of above methods there are some mutual problems like holes, blocking, etc. which comes from the structure of the motion vector processing in our proposed method by using a novel Image registration technique we overcame the problems both in using a medical approach to solve the problem based on the medium and the structural problems which made improvement in some video quality factors such as blocking and blurring and other video quality metrics. This method can be a basis and commence for researchers at using image registration to up convert frames on ultrasound images to improve further video quality parameters
  9. Keywords:
  10. Frame Rate Up Conversion ; Motion Compensation ; Motion Vector ; Image Registration ; Bluriness ; Blockiness

 Digital Object List

 Bookmark

No TOC