Lossy channel video blur avoidance
Blurriness in a video image transmitted over a lossy network can be reduced by filtering the decoder output to determine whether updating of the reference frame based on the current video data is indicated. An update may be appropriate if the decoder output is sharper than the contents of the display buffer after the last update. Additionally, an update may be appropriate if there was a significant change to the underlying image, and thus to the encoder's input. Whether such a change has occurred can be determined by analyzing the prediction mode, motion vectors, and pixel data of the output image.
1. A method for reducing blurriness of video data, the method comprising:
receiving video data at a decoder;
decoding a picture unit derived from the video data, the decoded picture unit corresponding to a picture unit stored in a display buffer;
determining, at a filter operatively coupled to the decoder, whether to update a display buffer operatively coupled to the decoder and a display device, based, at least in part, on a comparison of a quantization step size of the decoded picture unit with a quantization step size of its corresponding picture unit; and
updating, by the filter, the display buffer with the decoded picture unit when the size of the quantization step of the decoded picture unit is smaller than or equal to the size of the quantization step of the corresponding picture unit stored in the display buffer.
2. The method of claim 1 , wherein the picture unit is a macroblock.
3. The method of claim 1 , wherein the filter is integrated with the decoder.
4. The method of claim 1 , further comprising:
determining, at the filter operatively coupled to the decoder, whether there is more than one motion vector associated with the decoded picture unit; and
updating the display buffer with the decoded picture unit upon such determination.
5. The method of claim 4 , wherein comparing a quantization step size of the decoded picture unit with a quantization step size of the corresponding picture unit occurs before determining whether there is more than one motion vector associated with the decoded picture unit.
6. The method of claim 5 , wherein determining whether to update the display buffer is further based on whether a prediction mode associated with the decoded picture unit has been received from the decoder.
7. The method of claim 6 , wherein determining whether to update the display buffer is further based on whether prediction mode information supplied by the decoder indicates that the decoded picture unit was intra coded in an inter frame.
8. The method of claim 5 , further comprising:
receiving motion vector data associated with the decoded picture unit from the decoder;
comparing the motion vector data to a threshold; and
updating the display buffer if the motion vector data exceeds the threshold.
9. The method of claim 5 , wherein determining whether to update the display buffer is further based on comparing pixel data of the two picture units or on comparing information derived from the pixel data of the two picture units.
10. The method of claim 9 , wherein comparing pixel data or information derived from the pixel data of the two picture units comprises:
computing a difference between the two picture units;
performing a discrete cosine transform on the computed difference, thereby generating a plurality of coefficients;
comparing the plurality of coefficients to a threshold; and
determining that the coefficients exceed the threshold.
11. The method of claim 10 , wherein the threshold is determined based on the quantization step size of the decoded picture unit.
12. The method of claim 10 , wherein the discrete cosine transform uses the same picture unit size as is used by the decoder.
13. The method of claim 9 , wherein comparing pixel data or information derived from the pixel data of the two picture units comprises:
computing a sum of absolute differences for each picture unit; and
determining if the sum of absolute differences exceeds a threshold.
14. The method of claim 13 , wherein the threshold is determined based on the quantization step size of the decoded picture unit.
15. A video system configured to reduce blurriness of video data, the video system comprising:
a decoder configured to receive video data; and
a filter operatively coupled to the decoder and configured to:
determine whether to update a display buffer of the video system with a decoded picture unit derived from the video data by, at least, comparing a quantization step size of the decoded picture unit with a quantization step size of its corresponding picture unit; and
update the display buffer with the decoded picture unit when the quantization step of the picture unit stored in the display buffer is larger than quantization step of the decoded picture unit.
16. The video system of claim 15 , wherein the picture unit is a macroblock.
17. The video system of claim 15 , wherein the filter is integrated with the decoder.
18. The video system of claim 15 , wherein determining whether to update the display buffer further comprises determining whether the decoded picture unit has been motion compensated.
19. The video system of claim 15 , wherein determining, by the filter, whether to update a display buffer of the video system with a decoded picture unit derived from the video data by the decoder is further based on at least one of:
receiving a prediction mode associated with the decoded picture unit from the decoder;
receiving motion vector data associated with the decoded picture unit from the decoder, comparing the motion vector data to a threshold, and determining that the motion vector data exceeds the threshold; and
comparing pixel data of the two picture units or comparing information derived from the pixel data.
20. The video system of claim 15 , wherein the video system is a multipoint control unit.