Date of Award


Publication Type

Doctoral Thesis

Degree Name



Applied sciences, Deinterlacing, Motion data, Video sequences


Professor Sid-Ahmed


Prof. Ahmadi




In this work an efficient motion adaptive deinterlacing method with considerable improvement in picture quality is proposed. A temporal deinterlacing method has a high performance in static images while a spatial method has a better performance in dynamic parts. In the proposed deinterlacing method, a motion adaptive interpolator combines the results of a spatial method and a temporal method based on motion activity level of video sequence.

A high performance and low complexity algorithm for motion detection is introduced. This algorithm uses five consecutive interlaced video fields for motion detection. It is able to capture a wide range of motions from slow to fast. The algorithm benefits from a hierarchal structure. It starts with detecting motion in large partitions of a given field. Depending on the detected motion activity level for that partition, the motion detection algorithm might recursively be applied to sub-blocks of the original partition. Two different low pass filters are used during the motion detection to increase the algorithm accuracy. The result of motion detection is then used in the proposed motion adaptive interpolator.

The performance of the proposed deinterlacing algorithm is compared to previous methods in the literature. Experimenting with several standard video sequences, the method proposed in this work shows excellent results for motion detection and deinterlacing performance.