Face segmentation for use in real-time video coding and variable blocking to improve video

Today, bandwidth or storage space is one of the most challenging issues for transferring image or video. This challenge has already been raised since computers have had low storage space because with the advancement of technology, video image accuracy and details are improved, and naturally in order to transfer video, more storage space or more bandwidth is needed. The growth of image quality led to an important challenge in transforming, because for the storage, creating this space is easier than when the video needs to be transformed on the transferring environment, either the cost of upgrading is high or it’s not even in our authority. For example, if a video needs to be recorded, itsstorage space can be easily upgraded if needed, but the video call on the Internet is limited to the country, city, or even the source and destination of the transfer, that it’s upgrading is not easy. That s why video compression creates a big challenge.With the growth of VoIP technology and Internet telephony, video conversations have found their place in all areas, including home, industrial, medical, etc. Since these conversations are Real Time, any reduction in bandwidth or increase in video volume may disrupt the conversation so that the conversation looks tedious. Therefore, in video conversations, after increasing bandwidth, the most important principle is compression of the image. In this paper, considering that in the conversation, the most important part is the face of the audience, after facial recognition, using the techniques of image processing, it is attempting to transfer it with smaller blocks to increase accuracy, and thebackground image with larger blocks.

Keywords:

Image compression , video coding , VoIP