METHOD FOR PROCESSING DIGITAL IMAGE DATA

Abstract

The present invention relates to a method for processing digital image data (ID), wherein the process of multi rate, multi resolution and/or multi scale signal processing is involved in order to realize a respective multi rate, multi resolution and/or multi scale sharpness enhancement with respect to said image data (ID).

Description

BRIEF DESCRIPTION OF THE DRAWINGS

The invention will now be explained based on preferred embodiments thereof and by taking reference to the accompanying and schematical figures.

FIG. 1 is a schematical block diagram for elucidating a prior art signal processing method.

FIG. 2 is a schematical block diagram for elucidating a first embodiment of the inventive method for processing digital image data and in particular the edge-based multi rate image enhancement.

FIG. 3 demonstrates by means of a graphical representation the change of a gradient of an edge according to a process of decimation according to the present invention.

FIG. 4 is a schematical block diagram for elucidating the general inventive concept for a iterative process behind the inventive method for processing image data.

FIG. 5 is a schematical block diagram for elucidating details of each iteration step of the process shown in FIG. 4 in more detail.

FIGS. 6A-C are schematical block diagrams for elucidating details of the structure of the iteration step shown in FIG. 5 for different embodiments of the inventive method for processing digital image data.

Claims

1. Method for processing digital image data, comprising: (a) a step (SI) of providing digital image data (ID) to be processed as input data (inpD), said digital image data (ID) being representative for at least one of an image (I), a sequence of images (SI) and a video (V),(b) a step (S2) of processing said input data (inpD), thereby generating processed image data (procID) as processed data (procD), said processed image data (procID) being representative for an respective one of a processed image (procI), a processed sequence of images (procSI) and a processed video (procV), said processed image (procI), a processed sequence of images (procSI) and a processed video (procV), having enhanced sharpness properties, and(c) a step (S3) of providing said process data (procD) as output data (outD),wherein said step (S2) of processing said input data (inpD) comprises a process of detecting and enhancing edges (DEE) and uses a multi rate signal processing, multi resolution signal processing and/or multi scale signal processing with respect to said input data (inpD) in order to realize the multi rate, multi resolution and/or multi scale sharpness enhancement with respect to said input data (inpD).

2. Method according to claim 1, wherein said step (S2) of processing said input data (inpD), said process of detecting and enhancing edges (DEE) and/or said edge based sharpness enhancement are realized by a process of decomposition (D) and reconstruction (R) of said input data (inpD).

3. Method according to any one of the preceding claims, wherein said step (S2) of processing said input data (inpD), said process of detecting and enhancing edges (DEE), said edge based sharpness enhancement and/or said processes of decomposition (D) and reconstruction (R) are realized by a Laplace pyramid decomposition and reconstruction scheme.

4. Method according to any one of the preceding claims, wherein said step (S2) of processing said input data (inpD), said process of detecting and enhancing edges (DEE), said edge based sharpness enhancement, said processes of decomposition (D) and reconstruction (R) and/or said Laplace pyramid decomposition and reconstruction scheme are based on or use a process of decimation (DEC) and a process of interpolation/integration (INT).

5. Method according to claim 4, wherein said process of decimation (DEC) is based on a multi rate signal processing, multi scale signal processing and/or multi resolution signal processing.

6. Method according to any one of the preceding claims 4 or 5, wherein said process of interpolation/integration (INT) is based on a multi rate signal processing, multi scale signal processing and/or multi resolution signal processing.

7. Method according to any one of the preceding claims 4 to 6, wherein said process of decimation (DEC) comprises sub-processes of:(d1) low pass filtering (L) and/or anti-alias filtering (L) and of(d2) down-sampling (↓),in particular in that given order.

8. Method according to any one of the preceding claims 4 to 7, wherein said process of interpolation/integration (INT) comprises sub-processes of:(i1) up-sampling (↑) and of(i2) low pass filtering (L) and/or anti-alias filtering (L),in particular in that given order.

9. Method according to any one of the preceding claims 4 to 8, wherein said process of decimation (DEC) and/or said process of interpolation/integration (INT) and in particular the respective sub-processes (d1), (d2), (i1), (i2) thereof are carried out in order to reduce high frequency components, noise components and/or respective variances thereof andin order to keep the useful signal components of respective intermediate signals essentially unchanged or to reduce said useful components of respective intermediate signals only by a comparable smaller amount or by a comparable small amount, or unchanged.

10. Method according to any one of the preceding claims 4 to 9, wherein said process of decimation (DEC) and/or said process of interpolation/integration (INT) and in particular the respective sub-processes (d1, i2) of low pass filtering (L) and/or of anti-alias filtering (L) are based on a windowing process, e.g. on a Hamming window.

11. Method according to any one of the preceding claims 4 to 10, wherein said process of decimation (DEC) and/or said process of interpolation/integration (TNT) and in particular the respective sub-processes (d1, i2) of low pass filtering (L) and/or of anti-alias filtering (L) are pre-estimated based on a transfer function (H) given by said low pass filter (L) and/or by said anti-alias filter (L) which is involved.

12. Method according to claim 11, wherein the respective transfer function (H) of the underlying filter (L) is used in order to define at least one of a change factor, a variance range and a variance tolerance range in order to decide whether an area or signal component of said signal to be classified is dominated by high frequency signal components or noise.

13. Method according to any one of the preceding claims, wherein said step (S2) of processing said input data (inpD), said process of detecting and enhancing edges (DEE), said edge based sharpness enhancement, said processes of decomposition (D) and reconstruction (R), said Laplace pyramid decomposition and reconstruction scheme and/or said process of decimation (DEC) and said process of interpolation/integration (TNT) are realized iteratively with a multiplicity of iteration steps (k=1, . . . , N).

14. Method according to claim 13, wherein in each of said iteration steps (k=1, . . . , N) a detail signal (detk, k=0, 1, . . . , N) is generated.

15. Method according to any one of the preceding claims 13 or 14, wherein said iteration and in particular a respective iteration stop condition thereof are based on respective threshold values and/or on respective threshold conditions, in particular in a predefined manner.

16. Method according to any one of the preceding claims 13 to 15, wherein in each iteration step (k=1, . . . , N) a respective decomposition step (Dk) and a respective reconstruction step (Rk) are performed within said process of decomposition and within said process of reconstruction, respectively.

17. Method according to any one of the preceding claims 13 to 16, wherein in each iteration step (k=1, . . . , N) a respective decomposition step (Dk) receives input data (ink) from and generated by an—in particular directly—preceding iteration step (k−1) or said input data (inpD) if no iteration step is preceding.

18. Method according to claim 17, wherein in each iteration step (k=1, . . . , N−1) a respective decomposition step (Dk) generates and provides input data (ink+1) for/to a decomposition step (Dk) of an—in particular directly—succeeding iteration step (k+1) or output data (outk) for/to a respective reconstruction step (Rk) of the same iteration step (k) if no iteration step is succeeding.

19. Method according to claim 18, wherein in each iteration step (k=1, . . . , N−1) a respective decomposition step (Dk) generates and provides said input data (ink+1) according to equations (1) and (2) inpk:=DEC(inpk−1)=↓(L(inpk−1)) for k=1, . . . , N   (1)inp0:=inpD,   (2)wherein inpD denotes the input data of the whole process, inpk denotes the input data for the k-th iteration step, DEC(•) denotes the decimation process (DEC), ↓(•) denotes the down sampling process (d2), and L(•) denotes the low pass filtering and/or anti-aliasing filtering process (d1).

20. Method according to any one of the preceding claims 13 or 19, wherein in each iteration step (k=1, . . . , N) a respective decomposition step (Dk) generates and provides detail data (detk) for/to a respective reconstruction step (Rk) of the same iteration step (k).

21. Method according to claim 20, wherein in each iteration step (k=1, . . . , N−1) a respective decomposition step (Dk) generates and provides said detail data (detk) according to one of the cases (a), (b) and (c) of equations (3)

22. Method according to any one of the preceding claims 13 to 21, wherein in each iteration step (k=1, . . . , N) a respective reconstruction step (Rk) receives detail data (detk) from and generated by a respective decomposition step (Dk) of the same iteration step and output data (outk) from and generated by a respective reconstruction step (Dk+1) of an—in particular directly—succeeding iteration step (k+1) or said input data (ink) of said decomposition step (Dk) of the same iteration step if no iteration step is succeeding.

23. Method according to any one of the preceding claims 13 to 22, wherein in each iteration step (k=1, . . . , N−1) a respective reconstruction step (Rk) generates and provides output data (outk−1) for a and to a reconstruction step (Rk−1) of an—in particular directly—preceding iteration step (k−1) or output data (outD) for the whole process if no iteration step is preceding.

24. Method according to claim 23, wherein in each iteration step (k=1, . . . , N−1) a respective reconstruction step (Rk) generates and provides and/or receives said output data (outk) according to one of the cases (a), (b) and (c) of equation (5) and according to equation (4) and (6):

25. Method according to any one of the preceding claims, wherein edge detection (DEE) is carried out from the original input signal (inpD) and the decimator (DEC) output on the decomposition side (D) of a multi rate signal processing.

26. Method according to any one of the preceding claims 1 to 24, wherein edge detection (DEE) is carried out from the output of the interpolator (INT) on the decomposition side (D) of a multi rate signal processing.

27. Method according to any one of the preceding claims 1 to 24, wherein edge detection (DEE) is carried out from the output of the interpolator (INT) on the reconstruction side (R) of a multi rate signal processing.

28. Method according to any one of the preceding claims, wherein the process of detecting and enhancing edges is based on a process of edge-based image sharpness enhancement.

29. Method according to any one of the preceding claims, wherein the process of detecting and enhancing edges is based on a process of non edge-based image sharpness enhancement.

30. Method according to any one of the preceding claims, wherein edge detection is carried out on the basis of multi-resolution signal processing, and then enhanced by an edge-based sharpness enhancement method.

31. Method according to any one of the preceding claims, wherein for edges that cannot be detected on a higher resolution level are detected on a lower resolution level, and therefore further enhanced.

32. Method according to any one of the preceding claims, wherein the edge detection on different resolution levels is done using different or the same edge threshold value.

33. Method according to any one of the preceding claims, wherein the edge enhancement amounts on different resolution levels are controlled so that different signal parts can be emphasized in a controlled manner.

34. System, apparatus, or device for processing digital image data, which is adapted and which comprises means for carrying out a method for processing digital image data according to any one of the preceding claims 1 to 33 and the steps thereof.

35. Computer program product, comprising computer program means which is adapted in order to carry out the method for processing digital image data according to any one of the preceding claims 1 to 33 and the steps thereof when it is carried out on a computer or a digital signal processing means.

36. Computer readable storage medium, comprising a computer program product according to claim 35.