What is image compression? Part 3
What is an image that humans feel “beautiful”?
Humans focus on what they want to see, not the whole. If you improve the image quality of the area of interest, you will feel that it is the “cleanest” image. Therefore, reduce the compression rate of the part you are paying attention to to improve image quality, and increase the compression rate of the part you are not paying attention to to reduce the amount of information. In this way, we perform a “sharp picture control” that changes the compression ratio in a well balanced way.
Color Bleed Prevention Technology
With conventional technology, the image will deteriorate if compression and restoration are repeated many times. For example, a phenomenon called “color bleed” is known in which the outline becomes unclear.
What does it mean that the image gets worse?
From the footage of the scene to our television, we will go through multiple relay points. Each time it passes through the relay point, the video is restored, compressed and transmitted. If you repeat compression and restoration many times, the color of the characters in the image will bleed.
First of all, why do you reset the tablet every time you go through the relay point?
It would be nice if it could be brought from the site to the TV in a compressed state at once, but for the images you see on the news, etc., you can switch between multiple images. The switching device is called a switch. To use the switcher, you cannot use the compressed video as is, but restore it before editing. A switch is a device included in each relay point.
Developed technology
There’s a reason video gets worse every time it’s compressed and restored across multiple broadcast points. Images taken in situ are first divided into “color components” and “gloss components”. The vertical resolution of that “color component” is cut in half and then the whole is compressed. (At this time, the “glow component” does not shrink. The human eye is sensitive to “glow”, so try not to retain it too much.) By reducing the vertical resolution by half, the color bleed gradually builds up with the conventional formula and, as it repeats, becomes visible to the human eye. Therefore, we have developed an original Fujitsu calculation method that does not cause color bleed even if compression and restoration are repeated many times.
Recognized as a global standard
The Society of Film and Television Engineers (Society of Film and Television Engineers), which sets world broadcast standards, has adopted this technology as a recommended technology. Not to mention the excellent color bleed prevention technology, there is another reason why it was adopted. It was decided that the device with this new technology and the existing device could be used together. If this technology cannot be used with older equipment, you will have to replace all equipment in several repeater stations at the same time. Not very realistic. This time, it was adopted because it was a technology that was easy for the company operating the repeater station because the equipment only needed to be replaced when possible.
Difference between “H.264” and “H.265” for those who want to know more
(On this page, “H.264 / MPEG-4 AVC” is written as “H.264” and “H.265 / MPEG-H HEVC” is written as “H.265”)
Difference in compression ratio
“H.264” achieves twice the compression speed of MPEG-2. “H.265” achieves a compression rate four times that of MPEG-2. In other words, “H.265” can achieve the same image quality with half the file size and bit rate of “H.264”.
Difference in compression time
For example, if you compress the same video data to “H.264” and “H.265”, “H.265” has a higher compression rate, so the compression will take longer, but it is expected that do gradually. abbreviated. I can do it.
Why is the compression ratio of “H.265” higher than that of “H.264”?
Block size optimization technology can increase the compression ratio. In the case of moving images, the entire screen is subdivided (locked) and only the modified information is sent. In “H.264”, only changes are sent in small blocks, but blocks with few changes and blocks with high complexity are sent in the same small blocks.
“H.265 optimizes the blocks according to the amount of change. That is, a block with few changes is a large block and a block with complex changes is a small block. Then the total amount of information can be reduced.
