H.265+ Encoding Technology: Hikvision
H.265+ Encoding Technology: Hikvision
H.265+ Encoding Technology: Hikvision
H.265+ Encoding
Technology
Halve Your Bandwidth and Storage
Enjoy the Ultra HD and Fluency
Table of Contents
1. Background ............................................................................................................................... 3
2. Key Technologies ...................................................................................................................... 3
2.1. Prediction Encoding ..................................................................................................................... 3
2.1.1. P-Frame Encoding ...................................................................................................................... 4
2.1.2. I-Frame and R-Frame Encoding ................................................................................................. 5
2.2. Noise Suppression ....................................................................................................................... 7
2.3. Long-Term Bitrate Control ........................................................................................................... 7
3. Bitrate Reduction Test Result................................................................................................... 8
3.1. Instant Bitrate of Different Scenes ............................................................................................. 9
3.2. 24-Hour File Size of Different Scenes ....................................................................................... 10
4. Summary .................................................................................................................................. 11
The moving objects may appear only several times in a substantial portion of
surveillance videos.
24 hours non-stop surveillance and the noise on the video has a relatively great
impact on image quality.
H.265+ is able to greatly decrease the bitrate of the surveillance video so as to reduce the
bandwidth and storage cost.
2. KEY TECHNOLOGIES
H.265+ improves the compression ratio based on three key technologies: prediction
encoding technology based on the background model, background noise suppression
technology, and long-term bitrate control technology.
Inter-frame prediction creates a prediction model from one or more previously encoded
video frames or fields using block-based motion compensation.
Intra-frame prediction means that the samples of a macroblock (processing unit) are
predicted by using only information of already transmitted macroblocks of the same
frame.
For different frames in the video stream, different encoding methods are adopted. I-frame
can be encoded separately and it adopts the intra-frame prediction encoding technology.
P-frame encoding relies on the encoding of I-frame or P-frame, and it adopts the inter-
For a surveillance video, the background information is usually stable. Thus, we can extract
a background frame as a reference frame to encode. The background frame should contain
as few moving objects as possible.
As shown in Figure 1, among the three frames, T0 and T1 are the encoded images. You can
take a background frame as a reference frame, and then encode T2 based on the
similarities and differences between T1 (reference frame) and T0 (background frame). Since
T0 contains fewer moving objects, it s a good choice as the background frame.
T0 T1 T2
We can extract the moving object from T1 and T2, as shown in Figure 2. The object moves B
to A, when encoding the T2 frame, the B area is the newly exposed area.
Moving direction
B A
Figure 2 The Object Moves from B to A
Example 1:
If we take only T1 as the reference frame, it will cost more bits when encoding the B area
since we cannot find an area in T1 that is similar to the B area.
Example 2:
However, if we take only the background frame (T0) as the reference frame, we can find a
block that is almost the same as the B area, while the A area is a new area. Then it will cost
more bits when encoding the A area.
Example 3:
Moreover, if we take the background frame and T1 as the reference frames, we can simply
find the areas that are similar with the A area and B area. Then the bit cost can be reduced
to the lowest value.
T0
T2
T1
Reference frames
In order to reduce the bitrate cost of the repetitive I-frame, H.265+ designs a prediction
encoding reference relationship (based on background model) shown as Figure 5.
I-Frame I-Frame
(Background frame) (Background frame)
P-Frame Same as the The frame adopts the inter-frame prediction encoding that
frame rate is based on the previous frame (P-frame or R-frame) and I-
frame.
With the R-frame, bitrate cost can be lowered with the playback user experience is guaranteed
as well. See the figure below for the R-frame encoding process. The moving objects, which are
marked with red rectangles are encoded with the intra-frame prediction encoding and have
a good quality. The background adopts the inter-frame prediction encoding.
Moving objects
Final image
As shown in Figure 7, the intelligent analysis algorithm extracts the background image and
the moving vehicle. The background image is encoded with high compression in order to
suppress the noise and to lower the bitrate.
Intelligent Bitrate
analysis
Highly
Image Compressed
Original Compressed
Reduced Data
Bitrate Bitrate
Bitrate
Max. Bitrate
Time
0 6 12 18 24
Bitrate
Average Bitrate
Time
0 6 12 18 24
For H.265 encoding with the bitrate configured as constant, the bitrate slightly varies near
the predefined Max. bitrate value. With H.265+ on, the average bitrate can be kept as the
half value of the Max. bitrate (take the office surveillance as the example, the actual bitrate
reduction rate may vary with different surveillance scenes) and the image quality can be
optimized since the H.265+ technology makes full use of every bit.
In the variable bitrate mode, the instant bitrate varies according to the scene busyness,
while the image quality is steady. When H.265+ is on, the bitrate change can be divided into
two cases.
o If the set average bitrate value is limited, the H.265+ encoding can provide a better
image quality with the limited bitrate.
o If the set average bitrate value is high for the monitoring scene, the actual average
bitrate can be lower than the predefined value, thus the storage can be saved.
1. The average decreased rate between H.264 and Hikvision H.265+ is 83.7 percent, and
the rate between H.265 and Hikvision H.265+ is 67.6 percent. Hikvision H.265+ can
greatly decrease the bitrate in the same scene.
2. The decreased rate falls as the number of moving objects in the scene increases.
60.00%
Decreased Rate
15
50.00%
2.3 40.00%
10
17.9 30.00%
5 20.00%
9.5
0.56 10.00%
3.3
0 0.00%
H.264 H.265 Hikvision H.265+
Scene 2: Intersection
Decreased Rate
25
50.00%
20
40.00%
15 10
30.00%
10 22.7
20.00%
5 11.1 2.5
10.00%
5
0 0.00%
H.264 H.265 Hikvision H.265+
Conclusion:
For the café monitoring, the decreased rate of the 24-hour file size between H.264 and
Hikvision H.265+ is 83.0 percent, and the rate between H.265 and Hikvision H.265+ is 67.3
percent.
For the intersection monitoring, the decreased rate of the 24-hour file size between H.264
and Hikvision H.265+ is 79.4 percent, and the rate between H.265 and Hikvision H.265+ is
64.5 percent.
Hikvision H.265+ can greatly decrease the file size in the all-day monitoring and decrease
the storage cost in the end.
4. SUMMARY
Hikvision H.265+ is an optimized encoding technology based on the standard H.265/HEVC
compression. With H.265+, the video quality is almost the same as that of H.265/HEVC but
with less transmission bandwidth and storage capacity required. It extends the application
of the ultra HD video in video surveillance, such as 8MP and 12MP devices.
Hikvision H.265+ meets the H.265/HEVC standard, and it is compatible with most of the
software/hardware supporting H.265. It plays a significant role in cutting the storage cost
and promoting the ultra-HD video popularity.