Motion video signal processing for recording or reproducing – Local trick play processing – With randomly accessible medium
Reexamination Certificate
1999-10-06
2004-02-24
Tran, Thai (Department: 2615)
Motion video signal processing for recording or reproducing
Local trick play processing
With randomly accessible medium
C386S349000, C375S240030
Reexamination Certificate
active
06697567
ABSTRACT:
FIELD OF THE INVENTION
The present invention relates to a dynamic image encoding apparatus for real-time recording a dynamic image in a storage medium of a fixed recording capacity at a variable bit rate. The invention more particularly relates to a dynamic image encoding apparatus capable of recording data of a dynamic image of a prescribed time within a prescribed capacity of the storage medium while maintaining a picture quality level.
BACKGROUND OF THE INVENTION
Along with the popularization of the multi-media technology in recent years, dynamic image encoding apparatuses for carrying out variable bandwidth compression encoding such as MPEG (moving picture experts group) encoding system or the like have been frequently used. Regarding this encoding system, MPEG1 is prescribed in ISO13818 and MPEG2 is prescribed in ISO11172, for example.
FIG. 22
is a block diagram showing a structure of a conventional dynamic image encoding apparatus. Referring to
FIG. 22
, at first, an input original image
1
according to the NTSC system or the like to be coded and recorded is input to an input control section
2
. Then, the input control section
2
time-filters and space-filters the input image to divide it into various kinds of pictures in the MPEG system including I picture (an intra-frame predicted image), P picture (a forward inter-frame predicted image) and B picture (a bi-directional inter-frame predicted image), rearranges the sequence of encoding and outputs input image data
3
divided into macro block units.
A movement detector
6
detects a movement of the P picture and the B picture based on the input image data
3
and inter-frame predicted image data
5
output from an image frame memory
4
. Regarding the P picture and the B picture, an inter-frame subtractor
7
subtracts movement-compensated inter-frame predicted image data
5
from the input image data
3
, and outputs a result to a DCT (a discrete cosine transformer)
9
as inter-frame differential data
8
. Regarding the I picture, the inter-frame subtractor
7
sets the inter-frame predicted image data
5
to zero and outputs the input image data
3
to the DCT
9
as inter-frame differential data
8
.
The DCT
9
discrete cosine transforms the inter-frame differential data and outputs a DCT output
10
to a quantizer
11
. The quantizer
11
quantizes the DCT output
10
and produces a quantizer output
12
. A variable length encoder
13
variable-length encodes the quantizer output
12
, and outputs resultant bandwidth-compressed data as a transmission buffer input
14
. On the other hand, in order to carry out a local decoding for generating inter-frame predicted images of subsequent image frames as an inter-frame predictive encoding according to differential pulse code modulation (DPCM) system, an inverse quantizer
15
inversely quantizes the quantizer output
12
and outputs a resultant inverse quantizer output
16
to an inverse discrete cosine transformer (an inverse DCT)
17
. The inverse DCT
17
inversely discrete-cosine transforms the inverse quantizer output
16
and outputs a resultant inverse DCT output
18
to an inter-frame adder
19
. The inter-frame adder
19
adds the inverse DCT output
18
and the inter-frame predicted image data
5
, and stores a result in the image frame memory
4
as inter-frame added data
20
. The inter-frame predicted image data of the I picture is zero.
A transmission buffer section
21
temporarily stores the transmission buffer input
14
, and transmits it to a channel adapter or a storage medium
26
of a prescribed bit rate as a transmission buffer output
22
at a fixed bit rate in synchronism with a clock of a fixed bit rate clock
27
. In this case, an encoding generated information quantity controller
109
controls a quantization step size
25
by using a transmission buffer status
23
from the transmission buffer section
21
.
FIG. 24
shows a relationship between a bit rate and the quantization step size
25
of each input image data
3
by characteristics. Referring to
FIG. 24
, when the quantization step size
25
is made larger in an images having the same characteristics, the quantization becomes coarse and the picture quality of the coded image deteriorates. However, a small value is generated in a quantized result, and information quantity generated by a variable-length encoding decreases, resulting into a reduction in the bit rate. On the contrary, when the quantization step size
25
is made smaller, the quantization becomes fine and the picture quality of the coded image improves. However, a large value is generated in a quantized result and information quantity generated by the variable-length encoding increases, resulting into an increase in a bit rate. The input image data
3
has a range of characteristics from a large image having a largest generated information quantity with a fine image and a rapid movement as shown by the line
3
a
in
FIG. 24
, to an image having a smallest generated information quantity with a simple image and little movement as shown by the line
3
c
in FIG.
24
. Usually, there exists, as a major portion, image input data of a standard generated information quantity as shown by the line
3
b
in FIG.
24
.
FIG. 23
is a block diagram showing a detailed structure of the transmission buffer section
21
and the encoding generated information quantity controller
109
. Referring to
FIG. 23
, a generated information quantity counter
101
counts at all times the transmission buffer input
14
and outputs a generated information quantity (a number of bits)
102
. A transmission information quantity counter
103
counts at all times the transmission buffer output
22
and outputs a transmission information quantity (a number of bits)
104
. A subtractor
105
subtracts the transmission information quantity
104
from the generated information quantity
102
and outputs a buffer stored information quantity
106
.
The generated information quantity
102
, the transmission information quantity
104
and the buffer stored information quantity
106
are input into the encoding generated information quantity controller
109
as a transmission buffer status
23
. The encoding generated information quantity controller
109
samples the input transmission buffer status
23
with registers
107
, and determines a quantization step size with CPU
108
based on this transmission buffer status
23
, and outputs a result as the quantization step size
25
to the quantizer
11
and the inverse quantizer
15
.
When the buffer capacity of the transmission buffer
21
a
is B, for example, encoding is carried out in a quantization step size Q
0
and information is stored until the buffer capacity reaches B/2. When the buffer capacity has reached B/2, transmission is started from this time at a fixed transmission bit rate R
0
. Thereafter, a differential obtained by subtracting the fixed transmission bit rate R
0
from the generated bit rate R of the transmission buffer input
14
is time-integrated, and the differential is increased or decreased as a buffer stored quantity.
The CPU
108
calculates the following expression by assuming that an average during one second of the buffer stored information quantities
106
for each frame to be A:
Q
(
n
+1)=
Q
(
n
)+(2
A/B
−1)·
Q
0
(1)
Then, the CPU
108
controls to make the stored information quantity of the transmission buffer section
21
come closer to B/2. In the above expression, Q (n+1) represents a quantization step size of a frame (n+1), Q (n) represents a quantization step size of a frame n, and Q
0
represents a constant. The CPU
108
carries out various controls such as the control of changing the ratio of the quantization step size
25
for each of the I picture, P picture and B picture.
FIG. 25
shows a relationship between a generated information quantity and a recording time according to the conventional dynamic image encoding apparatus. In
FIG. 25
, although the amount of changes are different depending on the time constant taken for
Onuaku Christopher
Renesas Technology Corp.
Tran Thai
LandOfFree
Dynamic image encoding apparatus does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Dynamic image encoding apparatus, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Dynamic image encoding apparatus will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-3354198