Methods and systems for computing the quality of an MPEG-2 video stream
View Patent ↗Novel “loss rate” metrics are presented that can be used to indicate the quality of a video stream. These metrics may be determined using a simplified method or by a more sophisticated method, the latter method applying different “weights” (i.e., adjustments) to error computations based on the type of frame involved (i.e., I, P, B). Both methods may produce a video quality score within a given range (e.g., in the range of 0 to 1) where one end of the range indicates a video stream of high quality (i.e., no frames are affected), and the other end indicates a video stream of low quality (i.e., all frames are affected).
1. A method for determining a total weighted loss rate of an MPEG-2 video stream comprising:
operating a device capable of capturing network packet transmissions to determine a weighted loss rate for an I frame using a weighted value that is based on an average frame distance between the I frame and a next I frame of a video stream in the time interval;
operating the device to determine a weighted loss rate for each P frame using a weighted value, where the weighted value for each P frame is based on an average frame distance between the P frame and the next I frame of the video stream;
operating the device to determine a weighted loss rate for each B frame using a weighted value, where the weighted value is associated with a number of P frames before the B frame; and
operating the device to sum the determined weighted loss rates of the I, P and B frames to arrive at a total, weighted loss rate.
2. The method as in claim 1 wherein the total weighted loss rate (λ m ) is given by
λ
m
=
1
T
m
·
[
D
i
·
L
m
i
+
(
1
-
L
m
i
T
m
i
)
·
D
p
_
·
∑
n
=
1
p
D
n
p
·
(
1
-
L
m
p
T
m
p
)
n
-
1
P
·
L
m
p
+
(
1
-
L
m
i
T
m
i
)
·
(
1
-
L
m
i
T
m
i
)
Pb
·
L
m
b
]
.
3. A device for determining a total weighted loss rate of an MPEG-2 video stream operable to:
determine a weighted loss rate for an I frame using a weighted value that is based on an average frame distance between the I frame and a next I frame of a video stream in the time interval;
determine a weighted loss rate for each P frame using a weighted value, where the weighted value for each P frame is based on an average frame distance between the P frame and the next I frame of the video stream;
determine a weighted loss rate for each B frame using a weighted value, where the weighted value is associated with a number of P frames before the B frame; and
sum the determined weighted loss rates of the I, P and B frames to arrive at a total, weighted loss rate.
4. The device as in claim 3 wherein the total weighted loss rate (λ m ) is given by
λ
m
=
1
T
m
·
[
D
i
·
L
m
i
+
(
1
-
L
m
i
T
m
i
)
·
D
p
_
·
∑
n
=
1
p
D
n
p
·
(
1
-
L
m
p
T
m
p
)
n
-
1
P
·
L
m
p
+
(
1
-
L
m
i
T
m
i
)
·
(
1
-
L
m
i
T
m
i
)
Pb
·
L
m
p
]
.
5. A method for determining a total weighted loss rate of a video stream comprising:
operating a device capable of capturing network packet transmissions to determine a weighted loss rate for an I frame using a weighted value that is based on an average frame distance between the I frame and a next I frame of a video stream in the time interval;
operating the device to determine a weighted loss rate for each P frame using a weighted value, where the weighted value for each P frame is based on an average frame distance between the P frame and the next I frame of the video stream;
operating the device to determine a weighted loss rate for each B frame using a weighted value, where the weighted value is associated with a number of P frames before the B frame; and
operating the device to sum the determined weighted loss rates of the I, P and B frames to arrive at a total, weighted loss rate.
6. The method as in claim 5 wherein the total weighted loss rate (λ m ) is given by
λ
m
=
1
T
m
·
[
D
i
·
L
m
i
+
(
1
-
L
m
i
T
m
i
)
·
D
p
_
·
∑
n
=
1
p
D
n
p
·
(
1
-
L
m
p
T
m
p
)
n
-
1
P
·
L
m
p
+
(
1
-
L
m
i
T
m
i
)
·
(
1
-
L
m
i
T
m
i
)
Pb
·
L
m
p
]
.