Systems and methods for triggering an out of focus alert
View Patent ↗In order to trigger an out of focus alert when the focus level of a video frame meets a focus criteria, a method is performed including the operations of: receiving a video frame, partitioning the video frame into a plurality of blocks, calculating an array of discrete cosign transformation (DCT) coefficients for at least one of the plurality of blocks using a DCT, classifying each of the at least one of the plurality of blocks based on the array of DCT coefficients for that block, calculating a focus level of the video frame from the block classifications, and triggering an out of focus alert if the focus level meets a focus criteria.
1. A computer system, comprising:
a storage system containing software; and
a processing system coupled to the storage system;
wherein the processing system is instructed by the software to:
receive a video frame;
bias the luminance of the video frame;
partition the video frame into a plurality of blocks;
calculate an array of discrete cosine transformation (DCT) coefficients for at least one of the plurality of blocks using a DCT;
generate block classifications by classifying at least one block of the at least one of the plurality of blocks as noisy based on the array of DCT coefficients for that block;
calculate a focus level of the video frame based at least in part on the block classifications; and
trigger an out of focus alert if the focus level meets a focus criteria.
2. The computer system of claim 1 ,
wherein each block is classified as one block type from the group of block types consisting of: dark, noisy, plain, unsharp, or sharp.
3. The computer system of claim 1 ,
wherein the focus level is calculated by dividing the number of sharp blocks by the sum of the number of unsharp blocks and sharp blocks.
4. The computer system of claim 1 ,
wherein the focus criteria is a focus threshold.
5. The computer system of claim 1 ,
wherein the focus criteria includes a persistence level.
6. The computer system of claim 1 ,
wherein the array of DCT coefficients for a block comprises 64 coefficients.
7. The computer system of claim 6 , wherein the processing system is further instructed by the software to:
calculate a dark value for the block; and
classify the block as a dark block if the dark value is less than a dark threshold.
8. The computer system of claim 7 ,
wherein the dark value is equal to dct[0], and
wherein dct[0:63] is the array of DCT coefficients, and dct[0] is the first element in the array of DCT coefficients.
9. The computer system of claim 8 ,
wherein the noise is equal to
∑
k
=
0
0
dct
[
k
]
∑
k
=
48
63
dct
[
k
]
,
and
wherein dct[0:63] is the array of DCT coefficients.
10. The computer system of claim 6 , wherein the processing system is further instructed by the software to:
scale the array of DCT coefficients by a weight factor contained in a weight array.
11. The computer system of claim 10 , wherein the processing system is further instructed by the software to:
calculate a scalar from the array of DTC coefficients;
calculate a lowScore from the array of DTC coefficients; and
calculate a highScore from the array of DTC coefficients.
12. The computer system of claim 11 ,
wherein calculate a scalar uses the formula:
scalar
=
min
(
16
,
∑
k
=
48
63
dct
[
k
]
+
2
)
;
wherein calculate a lowScore uses the formula:
lowScore
=
(
∑
k
=
1
5
[
dct
[
k
]
*
weight
[
k
]
scalar
]
)
*
scalar
;
wherein calculate a highScore uses the formula:
highScore
=
(
∑
k
=
6
63
[
dct
[
k
]
*
weight
[
k
]
scalar
]
)
*
scalar
;
and
wherein dct[0:63] is the array of DCT coefficients, and weight[0:63] is the weight array.
13. The computer system of claim 11 , wherein the processing system is further instructed by the software to:
classify the block as a plain block if lowScore is less than a plain threshold.
14. The computer system of claim 11 , wherein the processing system is further instructed by the software to:
classify the block as an unsharp block if lowScore is less than highScore.
15. The computer system of claim 11 , wherein the processing system is further instructed by the software to:
classify the block as a sharp block if:
lowScore is greater than or equal to highScore,
lowScore is greater than or equal to a plain threshold,
the dark value is greater than or equal to a dark threshold, and
the noise value is greater than or equal to a noise threshold.
16. The computer system of claim 1 , wherein the processing system is further instructed by the software to:
mask the video frame.
17. A non-transitory computer-readable medium of instructions for triggering an out of focus alert in a computer system, the instructions comprising:
receiving a video frame;
biasing the luminance of the video frame;
partitioning the video frame into a plurality of blocks;
calculating an array of discrete cosine transformation (DCT) coefficients for at least one of the plurality of blocks using a DCT;
generating block classifications by classifying at least one block of the at least one of the plurality of blocks as noisy based on the array of DCT coefficients for that block;
calculating a focus level of the video frame based at least in part on the block classifications; and
triggering an out of focus alert if the focus level meets a focus criteria.
18. The non-transitory computer-readable medium of instructions of claim 17 ,
wherein each block is classified as one block type from the group of block types consisting of: dark, noisy, plain, unsharp, or sharp.
19. The non-transitory computer-readable medium of instructions of claim 17 ,
wherein the focus level is calculated by dividing the number of sharp blocks by the sum of the number of unsharp blocks and sharp blocks.
20. The non-transitory computer-readable medium of instructions of claim 17 ,
wherein the focus criteria is a focus threshold.