Methods and systems for image processing

1. An image processing method implemented on at least one machine each of which has at least one processor and storage, the method comprising:

acquiring a target image, the target image including a plurality of elements, an element corresponding to a pixel or a voxel, the target image being generated by an imaging device;

decomposing, by a first decomposition algorithm, the target image into L layers, each layer of the L layers including a low frequency sub-image and a high frequency sub-image, L≥1;

decomposing, by a second decomposition algorithm, the target image into L′+N layers, each layer of the L′+N layers including a low frequency sub-image and a high frequency sub-image, L′≥1, and N≥1, the second decomposition algorithm being different from the first decomposition algorithm;

transforming at least one layer, the at least one layer including the Lth layer generated by the first decomposition algorithm or the L′th layer generated by the second decomposition algorithm;

updating the low frequency sub-image of the Lth layer generated by the first decomposition algorithm by replacing the low frequency sub-image of the Lth layer generated by the first decomposition algorithm with the low frequency sub-image of the L′th layer generated by the second decomposition algorithm; and

reconstructing, based on the high frequency sub-images of the L layers generated by the first decomposition algorithm, the updated low frequency sub-image of the Lth layer generated by the first decomposition algorithm, and the at least one transformed layer, a composite image.

2. The method of claim 1 , the acquiring a target image comprising:

acquiring an initial image;

extracting, based on the initial image, a region of interest (ROI);

extracting, based on the initial image, an ROI edge; and

determining, based on the ROI and the ROI edge, an ROI image as the target image.

3. The method of claim 2 , the extracting an ROI edge comprising:

denoising the initial image;

pre-processing, based on a gradient transform, the denoised initial image; and

detecting, based on the denoised initial image, the ROI edge.

4. The method of claim 1 , a low frequency sub-image of the at least one layer including a predetermined region including a plurality of gray levels, and the transforming the at least one layer comprising:

transforming the plurality of gray levels of the predetermined region.

5. The method of claim 4 , the transforming the plurality of gray levels of the predetermined region comprising:

determining a reference edge in the low frequency sub-image of the at least one layer;

determining, based on the low frequency sub-image of the at least one layer, a characteristic curve, the characteristic curve illustrating the relationship between a distance and a gray level corresponding to the distance, wherein the distance is a distance between a first element in the low frequency sub-image and a second element in the reference edge, and the first element corresponds to the second element, wherein the gray level corresponding to a distance is determined based on the plurality of gray levels of the predetermined region;

determining, based on the characteristic curve, a transformation curve, the transformation curve illustrating the relationship between a gray level corresponding to a distance before transformation and the corresponding gray level after transformation; and

updating, based on the transformation curve, the plurality of gray levels of the predetermined region.

6. The method of claim 5 , the determining a transformation curve comprising:

dividing the characteristic curve into M characteristic curve segments;

determining, based on the M characteristic curve segments, M transformation curve segments, wherein a characteristic curve segment corresponds to a transformation curve segment; and

generating, based on the M transformation curve segments, the transformation curve.

7. The method of claim 6 , the determining M transformation curve segments comprising:

for an xth transformation curve segment of the M transformation curve segments,

calculating a slope of the xth transformation curve segment based on a gray level of a predetermined point in the characteristic curve, a gray level of the starting point of an xth characteristic curve segment, and a gray level of the end point of the xth characteristic curve segment, the xth characteristic curve segment corresponding to the xth transformation curve segment, wherein x is an integer, 1≤x≤M;

determining a gray level of the starting point in the xth transformation curve segment, comprising:

if x=1, designating the gray level of the starting point in the xth characteristic curve segment as the gray level of the starting point in the xth transformation curve segment; and

if 1<x≤M, determining the gray level of the starting point in the xth transformation curve segment based on the gray level of the starting point of the (x−1)th transformation curve segment and a gray level variation of the (x−1)th characteristic curve segment.

8. The method of claim 6 further comprising:

determining a gray level range of the characteristic curve, wherein the gray level range is a range within which at least one gray level is to be transformed, and the gray level range corresponds to a portion of the characteristic curve; and

designating the maximum value or minimum value of the gray level range as the gray level of the predetermined point in the characteristic curve.

9. The method of claim 1 further comprising:

enhancing the high frequency sub-images of the L layers generated by the first decomposition algorithm.

10. The method of claim 1 , the reconstructing a composite image comprising:

for each of a plurality of iterations,

up-sampling the low frequency sub-image of the (L−i)th layer;

updating, based on the up-sampled low frequency sub-image of the (L−i)th layer and the high frequency sub-image of the (L−i)th layer, the low frequency sub-image of the (L−i−1)th layer, 0≤i≤L−1; and

reconstructing, based on the updated low frequency sub-image of the first layer and the high frequency sub-image of the first layer, the composite image.

11. The method of claim 10 , the up-sampling the low frequency sub-image of the (L−i)th layer comprising:

up-sampling, based on a bilinear interpolation or a cubic interpolation, the low frequency sub-image of the (L−i)th layer.

12. The method of claim 1 further comprising:

updating the low frequency sub-image of the L′th layer generated by the second decomposition algorithm based on the low frequency sub-image of the (L′+N)th layer generated by the second decomposition algorithm and the high frequency sub-images of the (L′+1)th layer through the (L′+N)th layer generated by the second decomposition algorithm.

13. The method of claim 1 , a high frequency sub-image of the at least one layer including a plurality of elements, and the transforming the at least one layer comprising:

generating a weight image for the high frequency sub-image, the weight image including a plurality of weights corresponding to the plurality of elements; and

updating, based on the weight image, the high frequency sub-image.

14. The method of claim 13 , the high frequency sub-image of the at least one layer including a first class of elements and a second class of elements, and the generating the weight image comprising:

determining a gray level range of the first class of elements in the high frequency sub-image;

determining, based on the gray level range of the first class of elements, a gray level range of the second class of elements in the high frequency sub-image;

mapping the gray level range of the first class of elements into [0, 1];

determining, based on the mapped gray level range of the first class of elements, weighting factors for the first class of elements;

mapping the gray level range of the second class of elements into (1, G], wherein G is a predetermined value;

determining, based on the mapped gray level range of the second class of elements, weighting factors for the second class of elements; and

generating, based on the weighting factors for the first class of elements and the weighting factors for the second class of elements, the weight image.

15. The method of claim 14 , the determining a gray level range of the first class of elements comprising:

determining, based on a gray level threshold, an initial gray level range of the first class of elements;

modifying the initial gray level range of the first class of elements; and

adjusting, based on the modified gray level range of the first class of elements, the initial gray level range of the first class of elements.

16. The method of claim 15 , the adjusting the initial gray level range of the first class of elements comprising:

calculating, based on the modified gray level range of the first class of elements, a first threshold; and

determining the gray level range of the first class of elements as [0, the first threshold].

17. The method of claim 13 , the updating the high frequency sub-image comprising:

multiplying the gray levels of the high frequency sub-image by that of the weight image.

18. The method of claim 1 , the transforming at least one layer comprising:

transforming a high frequency sub-image or a low frequency sub-image of the at least one layer by at least one of linear/nonlinear enhancement or denoising.

19. A non-transitory computer readable medium comprising executable instructions that, when executed by at least one processor, cause the at least one processor to effectuate a method comprising:

acquiring a target image, the target image including a plurality of elements, an element corresponding to a pixel or a voxel, the target image being generated by an imaging device;

decomposing, by a first decomposition algorithm, the target image into L layers, each layer of the L layers including a low frequency sub-image and a high frequency sub-image, L≥1;

decomposing, by a second decomposition algorithm, the target image into L′+N layers, each layer of the L′+N layers including a low frequency sub-image and a high frequency sub-image, L′≥1, and N≥1, the second decomposition algorithm being different from the first decomposition algorithm;

transforming at least one layer, the at least one layer including the Lth layer generated by the first decomposition algorithm or the L′th layer generated by the second decomposition algorithm;

updating the low frequency sub-image of the Lth layer generated by the first decomposition algorithm by replacing the low frequency sub-image of the Lth layer generated by the first decomposition algorithm with the low frequency sub-image of the L′th layer generated by the second decomposition algorithm; and

reconstructing, based on the high frequency sub-images of the L layers generated by the first decomposition algorithm, the updated low frequency sub-image of the Lth layer generated by the first decomposition algorithm, and the at least one transformed layer, a composite image.

20. A system comprising:

at least one processor, and

a storage configured to store instructions, the instructions, when executed by the at least one processor, causing the system to effectuate a method comprising:

acquiring a target image, the target image including a plurality of elements, an element corresponding to a pixel or a voxel, the target image being generated by an imaging device;

decomposing, by a first decomposition algorithm, the target image into L layers, each layer of the L layers including a low frequency sub-image and a high frequency sub-image, L≥1;

decomposing, by a second decomposition algorithm, the target image into L′+N layers, each layer of the L′+N layers including a low frequency sub-image and a high frequency sub-image, L′≥1, and N≥1, the second decomposition algorithm being different from the first decomposition algorithm;

transforming at least one layer, the at least one layer including the Lth layer generated by the first decomposition algorithm or the L′th layer generated by the second decomposition algorithm;

updating the low frequency sub-image of the Lth layer generated by the first decomposition algorithm by replacing the low frequency sub-image of the Lth layer generated by the first decomposition algorithm with the low frequency sub-image of the L′th layer generated by the second decomposition algorithm; and

reconstructing, based on the high frequency sub-images of the L layers generated by the first decomposition algorithm, the updated low frequency sub-image of the Lth layer generated by the first decomposition algorithm, and the at least one transformed layer, a composite image.