Image analysis – Image enhancement or restoration – Edge or contour enhancement
Reexamination Certificate
2000-06-29
2003-11-04
Mehta, Bhavesh M. (Department: 2625)
Image analysis
Image enhancement or restoration
Edge or contour enhancement
C382S190000, C382S194000, C382S199000, C382S220000, C382S243000, C382S250000, C382S272000, C382S273000, C382S274000, C382S275000, C348S252000, C358S001900
Reexamination Certificate
active
06643410
ABSTRACT:
FIELD OF THE INVENTION
This invention relates to the field of digital image processing, and more particularly to methods of determining the extent of blocking artifacts in a digital image.
BACKGROUND OF THE INVENTION
Digital images contain enormous amounts of data. Storage of such data on digital media is generally expensive, and the transmission of digital images requires either a large bandwidth or a long period of time. Many algorithms have been developed to compress image data by removing visually redundant information from the image. Discrete-cosine-transform-based (DCT-based) compression has been the most popular among existing techniques. The DCT has very good energy compaction and data decorrelation properties. Moreover, the DCT can be computed using fast algorithms and efficiently implemented using very large scale integration (VLSI) techniques. In DCT-based coding, an image is partitioned into small square blocks (typically 8×8) and the DCT is computed over these blocks to remove the local spatial correlation. In order to achieve high compression, quantization of the DCT coefficients is then performed. Quantization is an irreversible process that causes loss of information and distortions in the decompressed image. After quantization, the redundancy in the data is further reduced using entropy coding. At the decoder end, the received data is decoded, dequantized, and reconstructed by the inverse DCT. In general, a typical 8-bit gray-level image can be coded with a compression ratio of up to 10:1 without noticeable artifacts.
However, at low bit rates the reconstructed images generally suffer from visually annoying artifacts as a result of very coarse quantization. One major artifact is the blocking effect, which appears as artificial block boundaries between adjacent blocks. At a low bit rate, each block is represented mainly by the first few low-frequency coefficients and, since each block is processed independently, no interblock correlation is accounted for in standard block DCT-based coding schemes. Therefore, discontinuity across the block boundary becomes noticeable.
There are many techniques developed to reduce the blocking effect. (e.g., H. C. Reeve III and J. S. Lim, “Reduction of blocking effect in image coding,”
ICASSP
, pp. 1212-1215, 1983). Since most of these techniques employ some kind of image filtering technique and thus reduce image sharpness to some extent, it is imperative that these techniques not be used on a “good image,” i.e., one that has not been compressed highly enough to exhibit the blocking artifacts. Therefore, it is necessary to develop a triage algorithm that measures the extent of blocking artifacts in a digital image.
Certain blocking artifact removal algorithms require prior knowledge of block boundary locations. However, since digital images compressed with a DCT-based technique can be further modified, for example, through cropping and zooming, the block boundary locations may have changed and become unknown at the time of blocking artifact removal. Thus, for those blocking artifact removal algorithms, there is a need for an automatic technique for detecting block boundary locations in digital images having blocking artifacts.
Because digital images may or may not have been compressed, and may have gone through different degrees of compression, the extent of the blocking artifacts varies from nonexistent to visually objectionable. Therefore, There is a need for determining the extent of the blocking artifacts in order to decide whether to perform artifact removing procedures. Furthermore, because virtually all artifact removal algorithms remove blocking artifacts at the expense of image detail, there is a need for controlling the optimal amount of filtering in order to strive for a good trade-off between artifact removal and image detail preservation.
A conventional method for determining the extent of the blocking artifacts involves taking the ratio between the size of the compressed JPEG file and the size of the image, which corresponds to the amount of uncompressed image data. This ratio is commonly referred to as the compression ratio. However, the compression ratio is not necessarily a good measure of the quality of the compressed image, nor the extent of the blocking artifacts. In general, at the same compression ratio, a busy image would look worse than a less busy image because busy images are harder to compress (therefore a busy image has lost more details than the less busy image). Therefore, there is a need to define a more perceptually accurate measure of the blocking artifacts.
SUMMARY OF THE INVENTION
The present invention is directed to overcoming one or more of the problems set forth above. Briefly summarized, according to one aspect of the present invention, a digital image processing method for determining the extent of blocking artifacts in a digital image, includes the steps of: forming a column difference image; averaging the values in the columns in the column difference image to produce a column difference array; computing the average of the values in the column difference array that are separated by one block width to produce a block averaged column difference array; locating the peak value in the block averaged column difference array; calculating the mean value of the block averaged column difference array excluding the peak value to produce a column base value; computing the ratio between the peak value and the base value to produce a column ratio; repeating the above-mentioned steps in the row direction to produce a row ratio; and employing the column and row ratios as a measure of the extent of blocking artifacts in the digital image.
ADVANTAGES
The present invention has the advantage that it measures the extent of blocking artifacts in a digital image based on image content. The present invention also has the advantage that it automatically detects block boundary locations even if a digital image has been modified by cropping and zooming.
REFERENCES:
patent: 5682440 (1997-10-01), Yukawa
patent: 5774101 (1998-06-01), Hirai et al.
patent: 5793428 (1998-08-01), Coelho
patent: 5867221 (1999-02-01), Pullen et al.
patent: 5949919 (1999-09-01), Chen
patent: 5956430 (1999-09-01), Kunitake et al.
patent: 6058209 (2000-05-01), Vaidyanathan et al.
patent: 6411741 (2002-06-01), Hamamura et al.
patent: 6512855 (2003-01-01), Delean
patent: 6532306 (2003-03-01), Boon et al.
patent: 6563537 (2003-05-01), Kawamura et al.
patent: 2001/0036314 (2001-11-01), Yamaguchi et al.
patent: 2002/0044689 (2002-04-01), Roustaei et al.
patent: 0 986 270 a2 (2000-03-01), None
patent: 1 001 608 (2000-05-01), None
“Measuring blocking artefacts using harmonic analysis” by K.T. Tan and M. Ghanbari. Electronic Letters, Aug. 5, 1999, vol. 35, No. 16.
“Reduction of Blocking Effect in Image Coding” by Howard C. Reeve III and Jae S. Lim. ICASSP, pp. 1212-1215, 1983.
Joshi Rajan L.
Luo Jiebo
Yu Qing
Choobin Barry
Eastman Kodak Company
Mehta Bhavesh M.
Woods David M.
LandOfFree
Method of determining the extent of blocking artifacts in a... does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Method of determining the extent of blocking artifacts in a..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Method of determining the extent of blocking artifacts in a... will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-3176792