A Joint Compression Scheme of Video Feature Descriptors and Visual Content

ABSTRACT:

High-productivity pressure of visual element descriptors has as of late risen as a functioning point because of the quickly expanding interest in versatile visual recovery over data transfer capacity restricted systems. In any case, transmitting just those component descriptors may to a great extent limit its application scale because of the absence of essential visual substance. To encourage the across the board of highlight descriptors, a half breed structure of together packing the component descriptors and visual substance is exceptionally attractive.

In this project, such a substance in addition to highlight coding plan is explored, expecting to shape the up and coming age of video pressure framework toward visual recovery, where the high-productivity coding of both element descriptors and visual substance can be accomplished by misusing the collaborations between one another. From one viewpoint, visual component descriptors can accomplish minimized and effective portrayal by taking preferences of the structure and movement data in the compacted video stream.

To enhance the recovery execution, a novel rate-precision streamlining procedure is proposed to precisely appraise the recovery execution corruption in highlight coding. Then again, the effectively packed component information can be used to additionally enhance the video coding productivity by applying highlight coordinating based relative movement pay. Broad reproductions have demonstrated that the proposed joint pressure structure can offer critical bitrate decrease in speaking to both component descriptors and video outlines, while at the same time keeping up the best in class visual recovery execution.

LEAVE A REPLY

Please enter your comment!
Please enter your name here