Strip noise removal method based on adaptive moment matching

Abstract

The invention belongs to the technical field of image denoising, and particularly relates to a strip noise removal method based on adaptive moment matching. The method comprises the following steps: S1, acquiring an original image containing a supersaturation point based on a linear array detector, calculating mean values and standard deviations of elements in each row in the original image, and sorting the elements in each row according to a sequence of the mean values from small to large to obtain an image sequence; s2, selecting a cut-off factor, and intercepting the image sequence based on the cut-off factor to obtain an effective reference line; s3, constructing a to-be-corrected image based on each effective reference row, and calculating a mean value and a standard deviation of each column of elements of the to-be-corrected image; and S4, obtaining a correction coefficient of each pixel of the linear array detector based on the mean value and the standard deviation of each column of elements of the to-be-corrected image, and correcting the original image based on the correction coefficient of each pixel of the linear array detector to obtain a corrected image. According to the method, the information polluted by the strip noise in the image can be efficiently recovered, the availability of the image is improved, and the continuity of image observation is ensured.

Description

TECHNICAL FIELD

[0001] The present application belongs to the technical field of image denoising, and particularly relates to a strip noise removal method based on adaptive rectangle matching. BACKGROUND

[0002] In the push-broom imaging process of a linear array detector, the photoelectric conversion characteristics of each pixel of the linear array sensor are difficult to achieve absolute consistency due to the non-uniformity difference of the detector element itself, the dark current fluctuation, external factors and the like, and thus typical strip noise is introduced in the imaging result. As a kind of noise under special imaging conditions, the strip noise has a clear difference in the form of expression from the point noise. The strip noise is distributed along the push-broom direction of the detector, and the gray value of the strip noise is overall brighter or darker than that of the adjacent normal pixel column, forming a bright-dark alternating'strip interference', which directly destroys the gray uniformity and detail integrity of the image.

[0003] In the prior art, the rectangle matching method can effectively remove the response difference between the detector elements, but when there is an over-saturated scene in the image, the strip noise is removed by using the rectangle matching method, and at the same time, strip-shaped false marks are generated in the column where the over-saturated scene is located and the adjacent area. SUMMARY

[0004] Therefore, the present application aims to provide a strip noise removal method based on adaptive rectangle matching, so as to solve the problem that the existing strip noise removal technology represented by the rectangle matching method cannot adapt to the special gray characteristics of the over-saturated scene in the image, and a unified statistical correction logic is used for all pixel columns. Such indiscriminate correction can destroy the gray transition rule between the over-saturated area and the surrounding normal area, resulting in strip-shaped false marks near the column where the over-saturated scene is located when the strip noise is removed, and introducing new image distortion. The strip noise removal method proposed in the present application removes the strip noise for the image containing the over-saturated scene point, correctly and efficiently restores the information of the image polluted by the strip noise, improves the usability of the image, and ensures the continuity of the image observation. The strip noise removal method proposed in the present application can remove the strip noise for the image containing the over-saturated scene point, correctly and efficiently restore the information of the image polluted by the strip noise, and thus improve the usability of the image and ensure the continuity of the image observation. To achieve the above-mentioned purpose, the technical scheme of the present application is as follows: A strip noise removal method based on adaptive rectangle matching, specifically comprising the following steps: S1: obtaining an original image containing an over-saturated point based on a linear array detector, calculating the mean value and standard deviation of each row element in the original image, sorting each row element in the order of the mean value from small to large to obtain an image sequence; S2: selecting a cutoff factor, and cutting the image sequence based on the cutoff factor to obtain effective reference rows; S3: Construct the image to be corrected based on each valid reference row, and calculate the mean and standard deviation of each column element of the image to be corrected; S4: Obtain the correction coefficients of each pixel of the linear array detector based on the mean and standard deviation of each column element of the image to be corrected, and correct the original image based on the correction coefficients of each pixel of the linear array detector to obtain the corrected image.

[0005] Furthermore, in step S2, the cutoff factor ranges from 0 to 1.

[0006] Furthermore, in step S2, if the cutoff factor is ε, the rule for truncating the image sequence based on the cutoff factor is: remove the F×ε row elements that are sorted first and the F×ε row elements that are sorted last, where F is the total number of rows in the image sequence.

[0007] Furthermore, the formula for calculating the correction of the original image based on the correction coefficients of each pixel of the linear array detector is as follows: ; in, σ r Here is the column standard deviation of the image to be corrected. μ r The column mean of the image to be corrected. σ i For the first image to be corrected i Standard deviation of column elements μ i The first image to be corrected i The mean of the column elements. X i For the linear array detector before calibration i The output value of the column element, Y i For the linear array detector after calibration, the first i The output value of the column element.

[0008] Compared with the prior art, the present invention can achieve the following beneficial effects: (1) The strip noise removal method based on adaptive moment matching described in this invention effectively eliminates strip noise caused by uneven sensor response in push-broom imaging of linear array detectors, while avoiding strip artifacts caused by oversaturation points. Compared with the problem of artifacts still existing after processing by a single moment matching algorithm, it achieves coordinated removal of noise and artifacts by adaptively filtering effective rows (removing abnormal rows containing oversaturation points) and combining column direction correction coefficient calculation, which significantly improves the image quality after processing and meets the requirements of high-quality imaging.

[0009] (2) The strip noise removal method based on adaptive moment matching provided by the present application is based on the row mean, standard deviation and overall statistical characteristics of the image itself, without external reference data or manual threshold setting. The ordered screening of the row sequence F improves the accuracy of effective row selection, avoids the interference of oversaturated pixels from the source, adapts to complex imaging scenes with oversaturated scene points, and provides an efficient and reliable strip noise removal scheme for oversaturated images of line array detectors, with high technical practicability. BRIEF DESCRIPTION OF DRAWINGS The accompanying drawings, which form a part of the present application, are intended to provide further understanding of the present application and are incorporated herein for illustrative purposes. The schematic embodiments of the present application and their descriptions are used to explain the present application, and do not constitute an improper limitation on the present application. In the drawings: Figure 1 The flowchart of the strip noise removal method based on adaptive moment matching described in the embodiments of the present application; Figure 2 The structural schematic diagram of the strip noise removal method based on adaptive moment matching described in the embodiments of the present application; Fig. 3(a) is the original image described in the embodiments of the present application; Fig. 3(b) is the image processed by the single moment matching algorithm described in the embodiments of the present application; Fig. 3(c) is the corrected image obtained after processing by the strip noise removal method based on adaptive moment matching described in the embodiments of the present application. DETAILED DESCRIPTION

[0010] In order to make the purpose, technical scheme and advantages of the present application clearer, the present application will be further described in detail below in combination with the drawings and specific embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application, and do not constitute a limitation on the present application.

[0011] It should be noted that the embodiments in the present application and the features in the embodiments can be combined with each other without conflict.

[0012] In the description of the present application, it should be understood that the terms "center", "longitudinal", "transverse", "upper", "lower", "front", "back", "left", "right", "vertical", "horizontal", "top", "bottom", "inner", "outer" and the like indicate the orientation or positional relationship based on the orientation or positional relationship shown in the drawings, and are only for the convenience of describing the present application and simplifying the description, and do not indicate or imply that the device or element referred to must have a particular orientation, be constructed and operated in a particular orientation, and therefore cannot be understood as a limitation of the present application. In addition, the terms "first", "second" and the like are only for descriptive purposes and cannot be understood as indicating or implying relative importance or implicitly indicating the number of technical features indicated. Therefore, the features defined with "first", "second" and the like can be explicitly or implicitly included one or more. In the description of the present application, unless otherwise stated, the meaning of "a plurality of" is two or more.

[0013] In the description of the present application, it should be noted that unless otherwise explicitly specified and limited, the terms "mounting", "connection", "connection" should be understood broadly, for example, it can be fixedly connected, or it can be detachably connected, or integrally connected; it can be mechanically connected, or it can be electrically connected; it can be directly connected, or it can be indirectly connected through an intermediate medium, or it can be connected inside two elements. For those skilled in the art, the specific meaning of the above terms in the present application can be understood through specific circumstances.

[0014] The present application will be described in detail below with reference to the accompanying drawings and in conjunction with the embodiments.

[0015] As Figure 1 shown, the present application proposes a strip noise removal method based on adaptive matrix matching, which specifically includes the following steps: S1: obtaining an original image containing oversaturated scenes based on a linear array detector, calculating the mean and standard deviation of each row element in the original image, sorting each row element in order of mean from small to large, and obtaining an image sequence; S2: selecting a cutoff factor, and cutting the image sequence based on the cutoff factor to obtain effective reference rows; S3: constructing a to-be-corrected image based on each effective reference row, calculating the mean and standard deviation of each column element of the to-be-corrected image; S4: obtaining the correction coefficient of each pixel of the linear array detector based on the mean and standard deviation of each column element of the to-be-corrected image, and correcting the original image based on the correction coefficient of each pixel of the linear array detector to obtain a corrected image.

[0016] It should be noted that the original image formed by the push-broom of the linear array detector has oversaturated points, and the presence of oversaturated points will generate strip noise. This invention proposes an adaptive moment matching method for strip noise removal, which removes strip noise from images containing oversaturated points based on the data characteristics of the original image itself. First, effective reference rows are adaptively selected based on the mean and variance of the original image. Then, the column direction gain and offset correction coefficient of the original image are calculated to remove the strip noise. This invention can effectively remove strip noise in the original image caused by factors such as uneven sensor response. At the same time, it can remove the image band artifacts caused by oversaturated pixels. This invention targets original images with oversaturated scene points and removes image strip noise based on an adaptive moment matching algorithm.

[0017] In some embodiments, in step S2, the cutoff factor ranges from 0 to 1.

[0018] In some embodiments, in step S2, if the cutoff factor is ε, the rule for cropping the image sequence based on the cutoff factor is: remove the F×ε row elements that are sorted first and the F×ε row elements that are sorted last, where F is the total number of rows in the image sequence.

[0019] Using the total number of rows in the image sequence as the base, remove the first (total number of rows in the image sequence × ε) rows and the last (total number of rows in the image sequence × ε) rows in the image sequence, and retain the rows in the middle interval as valid reference rows, thereby eliminating abnormal rows containing oversaturation points.

[0020] In some embodiments, the formula for calculating the correction of the original image based on the correction coefficients of each pixel of the linear array detector is as follows: ; in, σ r Here is the column standard deviation of the image to be corrected. μ r The column mean of the image to be corrected. σ i The first image to be corrected i The standard deviation of the column elements μ i For the first image to be corrected i The mean of the column elements. X i For the linear array detector before calibration i The output value of the column element, Y i For the linear array detector after calibration, the first i The output value of the column element.

[0021] Example like Figure 2As shown, the present application proposes a strip noise removal method based on adaptive moment matching. Based on the original image itself feature information, the image strip noise removal is realized. Specifically: 1. Adaptive selection of effective reference row.

[0022] The mean and standard deviation of each row element in the original image are calculated, the rows in the original image are sorted in the order of mean from small to large based on the row direction, and the image sequence is obtained. The mean and standard deviation of the row element with oversaturation point will be significantly higher than those of other row elements. For this, a cutoff factor ε (ε∈(0,1)) is selected, and the row elements in the interval of ε~1-ε are intercepted, that is, after removing the row elements containing oversaturation points (scenery), the mean and standard deviation of the image to be corrected are calculated.

[0023] 2. Column direction correction coefficient calculation.

[0024] The mean and standard deviation of the column direction of the image to be corrected are calculated, the mean and standard deviation of each column element of the image to be corrected are calculated according to the moment matching theory, the correction coefficient of each pixel of the detector is calculated according to the following formula, the correction of the original image is realized, and the corrected image is obtained.

[0025] ; In the formula, σ r is the column standard deviation of the image to be corrected, μ r is the column mean of the image to be corrected, σ i is the standard deviation of the first i column element of the image to be corrected, μ i is the mean of the first i column element of the image to be corrected, X i is the output value of the first i column element of the linear array detector before correction, Y i is the output value of the first i column element of the linear array detector after correction Fig. 3(a) is an original image, and it can be clearly observed that there are both oversaturation points and strip noises in the image; Fig. 3(b) is the processing result of the single matrix matching algorithm, and since the algorithm does not design an adaptive solution for the oversaturation points, when the image has oversaturation points, the corresponding column will generate obvious banding artifacts in the strip noise removal process, which not only interferes with the removal effect of the strip noise, but also cannot meet the actual demand of high-quality imaging; Fig. 3(c) is the processing result of the present application, and it can be seen that the present application not only effectively eliminates the strip noise in the original image caused by factors such as non-uniform sensor response, but also successfully avoids the negative impact of the oversaturation points, without generating any banding artifacts, and the imaging quality is significantly better than that of the single matrix matching algorithm. In summary, the present application further improves the accuracy of selecting effective reference lines by ordering and screening the row sequence, and provides a more efficient and reliable strip noise removal scheme for the oversaturation points generated by the push-broom imaging of the linear array detector, and the technical effect is fully verified by the image comparison. It should be understood that various forms of flow shown above can be reordered, added, or deleted steps. For example, each step recorded in the present disclosure can be executed in parallel, sequentially, or in different order, as long as the desired results of the technical solutions of the present disclosure can be achieved, which is not limited herein.

[0026] The above specific embodiments do not constitute a limitation on the scope of protection of the present application. Those skilled in the art should understand that various modifications, combinations, sub-combinations and substitutions can be made according to design requirements and other factors. Any modifications, equivalent replacements and improvements made within the spirit and principles of the present application shall be included in the scope of protection of the present application.

Claims

1. A strip noise removal method based on adaptive moment matching, characterized by: Specifically comprising the following steps: S1: obtaining an original image containing a supersaturation point based on a linear array detector, calculating the mean and standard deviation of each row element in the original image, sorting each row element in order of the mean from small to large to obtain an image sequence; S2: selecting a cutoff factor, and cutting the image sequence based on the cutoff factor to obtain effective reference rows; S3: constructing a to-be-corrected image based on each effective reference row, and calculating the mean and standard deviation of each column element of the to-be-corrected image; S4: obtaining the correction coefficient of each pixel of the linear array detector based on the mean and standard deviation of each column element of the to-be-corrected image, and correcting the original image based on the correction coefficient of each pixel of the linear array detector to obtain a corrected image.

2. The strip noise removal method based on adaptive moment matching according to claim 1, characterized in that: In step S2, the value range of the cutoff factor is 0-1.

3. The adaptive moment-matching based strip noise removal method of claim 1, wherein: In step S2, if the cutoff factor is ε, the rule for cutting the image sequence based on the cutoff factor is to remove the first F×ε row elements and the last F×ε row elements in the order, and F is the total number of rows of the image sequence.

4. The adaptive moment-matching based strip noise removal method of claim 1, wherein: The calculation formula for correcting the original image based on the correction coefficient of each pixel of the linear array detector is: ； wherein, σ r is a column standard deviation of the image to be corrected, μ r is a column mean of the image to be corrected, σ i is a standard deviation of the first i column element of the image to be corrected, μ i is a mean of the first i column element of the image to be corrected, X i is an output value of the first i column element of the line array detector before correction, Y i is an output value of the first i column element of the line array detector after correction.

Images