CN113643192A - Fuzzy function processing method and device for imaging system, image acquisition equipment and storage medium - Google Patents
Fuzzy function processing method and device for imaging system, image acquisition equipment and storage medium Download PDFInfo
- Publication number
- CN113643192A CN113643192A CN202010395146.6A CN202010395146A CN113643192A CN 113643192 A CN113643192 A CN 113643192A CN 202010395146 A CN202010395146 A CN 202010395146A CN 113643192 A CN113643192 A CN 113643192A
- Authority
- CN
- China
- Prior art keywords
- processed
- image
- iteration
- blur function
- function
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 238000003384 imaging method Methods 0.000 title claims abstract description 44
- 238000003672 processing method Methods 0.000 title claims abstract description 25
- 238000000034 method Methods 0.000 claims abstract description 17
- 238000012545 processing Methods 0.000 claims description 35
- 239000011159 matrix material Substances 0.000 claims description 14
- 238000010606 normalization Methods 0.000 claims description 4
- 230000002349 favourable effect Effects 0.000 abstract description 3
- 238000012805 post-processing Methods 0.000 abstract description 3
- 230000006870 function Effects 0.000 description 180
- 238000010586 diagram Methods 0.000 description 8
- 230000000694 effects Effects 0.000 description 7
- 230000003287 optical effect Effects 0.000 description 6
- 230000009286 beneficial effect Effects 0.000 description 3
- 238000005516 engineering process Methods 0.000 description 3
- 238000012804 iterative process Methods 0.000 description 2
- 230000004044 response Effects 0.000 description 2
- 238000007476 Maximum Likelihood Methods 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 238000002474 experimental method Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 230000002035 prolonged effect Effects 0.000 description 1
- 230000001052 transient effect Effects 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T5/00—Image enhancement or restoration
- G06T5/73—Deblurring; Sharpening
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/10—Image acquisition modality
- G06T2207/10004—Still image; Photographic image
Landscapes
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Image Processing (AREA)
- Studio Devices (AREA)
Abstract
A fuzzy function processing method and device for an imaging system, an image acquisition device and a storage medium, wherein the method comprises the following steps: acquiring a blurred image, and determining an initial blurring function and an initial clear image; iteratively operating based on the blurred image, the blur function and the sharp image to obtain a processed sharp image and a processed blur function; and when the iteration result meets the stop condition, stopping iteration, and determining the processed fuzzy function obtained by the last iteration as the optimal fuzzy function. By the scheme of the invention, the fuzzy function caused by the movement of the collected object can be obtained, and the method is favorable for post processing to obtain a clear image.
Description
Technical Field
The invention relates to the technical field of image processing, in particular to a fuzzy function processing method and device for an imaging system, image acquisition equipment and a storage medium.
Background
In an image capturing operation such as photographing, movement of a subject causes imaging blur. For example, in a fingerprint recognition scenario, if a finger moves while the sensor is imaging, the image received by the sensor may be blurred due to the finger moving, and thus the fingerprint may not be recognized effectively.
In order to solve the imaging blurring problem, the prior art generally adopts a solution of re-shooting a clear image, but the shooting period is prolonged. For example, in a fingerprint unlocking scenario, when a finger continuously moves and a fingerprint needs to be continuously re-acquired, the user experience may be affected by an excessively long unlocking time.
Disclosure of Invention
The invention solves the technical problem of how to obtain a fuzzy function caused by the movement of an acquired object so as to obtain a clear image through post-processing.
To solve the above technical problem, an embodiment of the present invention provides a blur function processing method for an imaging system, including: acquiring a blurred image, and determining an initial blurring function and an initial clear image; iteratively calculating based on the blurred image, the blur function and the sharp image to obtain a processed sharp image and a processed blur function, wherein the blur function in the first iteration is the initial blur function, the sharp image in the first iteration is the initial sharp image, and the processed sharp image and the processed blur function obtained in the previous iteration from the second iteration are used as the sharp image and the blur function for calculating in the next iteration; and when the iteration result meets the stop condition, stopping iteration, and determining the processed fuzzy function obtained by the last iteration as the optimal fuzzy function.
Optionally, the stop condition at least includes: the similarity of the processed clear images obtained by the two iterations is greater than a preset threshold value.
Optionally, the fuzzy function is represented in a matrix form, and the determining the initial fuzzy function includes: assigning a random value to each element in the matrix to obtain the initial blur function.
Optionally, a value range of the random value is 0 to 1.
Optionally, the number of rows and the number of columns of the matrix are determined according to the size of the light source of the imaging system.
Optionally, the determining the initial sharp image includes: determining the blurred image as the initial sharp image.
Optionally, the iteratively operating based on the blurred image, the blur function, and the sharp image to obtain the processed sharp image and the processed blur function includes: and iteratively performing deconvolution operation based on the blurred image, the blur function and the sharp image to obtain a processed sharp image and a processed blur function.
Optionally, the iteratively performing deconvolution operation based on the blurred image, the blur function, and the sharp image to obtain the processed sharp image and the processed blur function includes: and calculating to obtain the processed clear image based on the following formula during each iteration:
wherein k represents the kth iterative operation, and k is more than or equal to 0; f. ofk+1(x) A processed clear image is obtained for the (k + 1) th iteration; f. ofk(x) A processed sharp image obtained for the kth iteration; g (x) is the blurred image; h isk(x) A processed fuzzy function obtained for the kth iteration; h isk(-x) is the inverse of the processed blur function obtained from the kth iteration; denotes a convolution operation; x represents a multiplication operation; and obtaining the processed fuzzy function based on the following formula during each iteration:
wherein h isk+1(x) A processed fuzzy function obtained for the (k + 1) th iteration; f. ofkThe result of the inversion of the processed sharp image obtained in the kth iteration is (-x).
Optionally, the processed clear image obtained by each iteration satisfies the following constraint: the gray scale of each pixel in the processed clear image is larger than zero.
Optionally, the processed fuzzy function obtained by each iteration satisfies the following constraint: the numerical value of each element in the processed fuzzy function is larger than zero; the processed fuzzy function meets the normalization condition; the number of rows and columns of the processed blur function remains unchanged.
Optionally, the stop condition further includes: the number of iterations reaches a preset number.
Optionally, when the iteration result satisfies the stop condition, stopping the iteration includes: and when the iteration times reach the preset times and the similarity between the processed clear image obtained by the last iteration and the processed clear image obtained by the previous iteration is smaller than the preset threshold, continuing the iteration until the similarity between the processed clear images obtained by the previous iteration and the processed clear image obtained by the next iteration is larger than the preset threshold.
Optionally, the fuzzy function processing method further includes: acquiring a to-be-processed blurred image; and deblurring and restoring the blurred image to be processed into a corresponding clear image based on the optimal blurring function.
Optionally, the deblurring and restoring the blurred image to be processed into a corresponding sharp image based on the preferred blur function includes: and restoring to obtain the clear image based on the following formula:
wherein,
to restore the resulting sharp image; g (x) is the blurred image to be processed; h (x) is the preferred blur function; *-1Representing a deconvolution operation.
In order to solve the above technical problem, an embodiment of the present invention further provides a blur function processing apparatus for an imaging system, including: the acquisition module is used for acquiring a blurred image and determining an initial blurring function and an initial sharp image; the iteration module is used for performing iterative operation on the basis of the blurred image, the blur function and the clear image to obtain a processed clear image and a processed blur function, wherein the blur function in the first iteration is the initial blur function, the clear image in the first iteration is the initial clear image, and the processed clear image and the processed blur function obtained in the previous iteration from the second iteration are used as the clear image and the blur function for performing operation in the next iteration; and the determining module stops iteration when the iteration result meets the stop condition and determines the processed fuzzy function obtained by the last iteration as the optimal fuzzy function.
Optionally, the stop condition at least includes: the similarity of the processed clear images obtained by the two iterations is greater than a preset threshold value.
To solve the above technical problem, an embodiment of the present invention further provides a storage medium having stored thereon computer instructions, where the computer instructions execute the steps of the above method when executed.
In order to solve the above technical problem, an embodiment of the present invention further provides an image capturing apparatus, including: the imaging system is used for acquiring a to-be-processed blurred image; a blur function processing module coupled to the imaging system, the blur function processing module for performing the above method to determine the preferred blur function; and the deblurring module is coupled with the fuzzy function processing module and the imaging system and is used for deblurring and restoring the blurred image to be processed into a corresponding clear image based on the optimal fuzzy function.
Compared with the prior art, the technical scheme of the embodiment of the invention has the following beneficial effects:
the embodiment of the invention provides a fuzzy function processing method for an imaging system, which comprises the following steps: acquiring a blurred image, and determining an initial blurring function and an initial clear image; iteratively calculating based on the blurred image, the blur function and the sharp image to obtain a processed sharp image and a processed blur function, wherein the blur function in the first iteration is the initial blur function, the sharp image in the first iteration is the initial sharp image, and the processed sharp image and the processed blur function obtained in the previous iteration from the second iteration are used as the sharp image and the blur function for calculating in the next iteration; and when the iteration result meets the stop condition, stopping iteration, and determining the processed fuzzy function obtained by the last iteration as the optimal fuzzy function.
Therefore, the scheme of the embodiment can find out the fuzzy function caused by the movement of the object based on an iterative algorithm, so as to be beneficial to the later processing, such as deblurring. Specifically, a sharp image is derived from a blurred image by an operation. Further, when the similarity of the clear images obtained by the current iteration and the later iteration is higher than a preset threshold value, it is indicated that the clear image obtained by the last iteration has a good restoration effect and tends to be stable, and the fuzzy function adopted in the operation of the iteration can be used as an optimal fuzzy function for deblurring the currently acquired fuzzy image.
Further, the stop condition includes at least: the similarity of the processed clear images obtained by the two iterations is greater than a preset threshold value. Compared with a scheme with a fixed iteration number as a stopping condition, the scheme with image similarity as the stopping condition is favorable for obtaining a better optimal fuzzy function. Specifically, since too many or too few iterations may affect the quality of the blur function and the degree of restoration of the blurred image, the deblurring effect of the image is not necessarily optimal for the optimal blur function obtained by processing with a fixed number of iterations as a stop condition. Based on the above, the scheme of the invention takes the image similarity as a stop condition, and determines the stop time according to the real-time processing effect in the iterative process, thereby obtaining a better optimal fuzzy function.
Further, an embodiment of the present invention further provides an image capturing apparatus, including: the imaging system is used for acquiring a to-be-processed blurred image; a blur function processing module coupled to the imaging system, the blur function processing module for performing the above method to determine the preferred blur function; and the deblurring module is coupled with the fuzzy function processing module and the imaging system and is used for deblurring and restoring the blurred image to be processed into a corresponding clear image based on the optimal fuzzy function.
By adopting the scheme of the embodiment, the acquired blurred image can be deblurred, and a clear image can be obtained without repeated acquisition. Therefore, the acquisition efficiency of the image acquisition equipment can be improved, and the acquisition time can be shortened.
Drawings
FIG. 1 is a flow chart of a blur function processing method for an imaging system according to an embodiment of the present invention;
FIG. 2 is a schematic diagram of a clear fingerprint image of a finger;
FIG. 3 is a schematic diagram of a blurred fingerprint image of a finger;
FIG. 4 is a theoretical diagram of a blur function caused by the movement of a finger;
FIG. 5 is a schematic illustration of a blur function obtained using the method of FIG. 1 for the image processing of FIG. 3;
FIG. 6 is a flow chart of a method of deblurring an image according to an embodiment of the present invention;
FIG. 7 is a schematic illustration of a blurred fingerprint image in a typical application scenario;
FIG. 8 is a sharp fingerprint image obtained using deblurring restoration based on the image shown in FIG. 7;
FIG. 9 is a diagram illustrating a blur function obtained by the method of the present embodiment with respect to the image processing shown in FIG. 7;
FIG. 10 is a schematic structural diagram of a blur function processing apparatus for an imaging system according to an embodiment of the present invention;
fig. 11 is a schematic structural diagram of an image deblurring apparatus according to an embodiment of the present invention.
Detailed Description
As background art, the prior art fails to provide a good solution when the captured object moves during image capture resulting in imaging blur.
Taking a fingerprint identification scene as an example, the conventional optical screen-type fingerprint identification technology utilizes a lens to image on a sensor, so that the module of the optical screen-type fingerprint identification technology needs elements such as a lens array, a light collimator, a spatial filter and the like, and the structure is complex, so that the defects of heavy module, small sensing range and the like are caused. The scheme of the embodiment adopts a new optical type underscreen fingerprint identification technology, is based on a total reflection imaging principle, and has the advantages of simple structure, light and thin module, low cost, easiness in realizing a large-area sensing range and the like.
However, when a fingerprint is captured, the finger may move when the sensor images, so that the image received by the sensor is blurred due to the movement, and thus, the fingerprint recognition may be difficult.
To solve the above technical problem, an embodiment of the present invention provides a blur function processing method for an imaging system, including: acquiring a blurred image, and determining an initial blurring function and an initial clear image; iteratively calculating based on the blurred image, the blur function and the sharp image to obtain a processed sharp image and a processed blur function, wherein the blur function in the first iteration is the initial blur function, the sharp image in the first iteration is the initial sharp image, and the processed sharp image and the processed blur function obtained in the previous iteration from the second iteration are used as the sharp image and the blur function for calculating in the next iteration; and when the iteration result meets the stop condition, stopping iteration, and determining the processed fuzzy function obtained by the last iteration as the optimal fuzzy function.
Therefore, the scheme of the embodiment can find out the fuzzy function caused by the movement of the object based on an iterative algorithm, so as to be beneficial to the later processing, such as deblurring. Specifically, a sharp image is derived from a blurred image by an operation. Further, when the similarity of the clear images obtained by the current iteration and the later iteration is higher than a preset threshold value, it is indicated that the clear image obtained by the last iteration has a good restoration effect and tends to be stable, and the fuzzy function adopted in the operation of the iteration can be used as an optimal fuzzy function for deblurring the currently acquired fuzzy image.
In order to make the aforementioned objects, features and advantages of the present invention comprehensible, embodiments accompanied with figures are described in detail below.
Fig. 1 is a flowchart of a blur function processing method for an imaging system according to an embodiment of the present invention.
The fuzzy function obtained by processing the scheme of the embodiment can be applied to an optical underscreen fingerprint identification scene. For example, the acquired blurred fingerprint image can be deblurred based on the blurring function to restore a clear fingerprint image from the blurred fingerprint image without repeatedly performing fingerprint acquisition operation. In practical application, the blur function may also be applied to other scenes in which post-processing is required for an acquired image that is blurred due to movement of an acquired object, so as to obtain a clear image based on blurred image restoration.
In an optical underscreen fingerprint recognition scenario, the imaging system may include: the fingerprint sensor comprises a light source component, a light-transmitting cover plate and a sensor component, wherein light rays emitted from the light source are totally reflected at the light-transmitting cover plate, and carry fingerprint information of a finger pressed on the light-transmitting cover plate to be incident to the sensor component. Thereby, the sensor part acquires a fingerprint image.
If the finger moves during imaging by the sensor unit, a blurred fingerprint image is acquired by the sensor unit. By adopting the scheme of the embodiment, the blurred fingerprint image can be deblurred based on the blurring function generated by the movement of the finger, so that a clear fingerprint image can be obtained.
The relationship among the sharp image, the blurred image and the blur function satisfies formula (1):
g(x)＝h(x)*f(x)+n (1)
wherein g (x) is the blurred image; h (x) is the blur function; (x) is the sharp image; is a convolution operation; n is system noise; (x) Is a matrix of pixels of the image.
The system noise n may be measured by the imaging system. For example, an image capturing operation is performed when no finger is placed on the light-transmitting cover plate to acquire an image without an object to be captured. The system noise n is typically used to characterize sensor noise, as well as ambient signal noise, etc.
The sharp image f (x) and the blurred image g (x) may be regarded as a matrix of pixels.
The blurring function h (x) may also be represented in matrix form.
In one embodiment, the blur Function h (x) is a Point Spread Function (PSF) describing the response of an imaging system to a Point source. Therefore, the result of convolving the point spread function h (x) with the sharp image f (x) (i.e. the eigen image of the object being acquired) is the image g (x) (i.e. the blurred image) actually acquired by the imaging system. The number of rows and columns of the matrix used here to characterize the blurring function h (x) may be determined according to the light source size of the imaging system. For example, the larger the area of the light source elements, the larger the number of rows and columns of the matrix. Besides the shape and size of the light source itself, the movement of the collected object also affects the blur function h (x), and the final blur function h (x) is formed by the superposition of two factors, namely the shape and size of the light source and the movement of the collected object.
In a specific implementation, referring to fig. 1, the fuzzy function processing method according to this embodiment may include the following steps:
step S101, acquiring a blurred image, and determining an initial blurring function and an initial sharp image;
step S102, iteratively calculating based on the blurred image, the blur function and the sharp image to obtain a processed sharp image and a processed blur function, wherein the blur function in the first iteration is the initial blur function, the sharp image in the first iteration is the initial sharp image, and the processed sharp image and the processed blur function obtained in the previous iteration from the second iteration are used as the sharp image and the blur function for calculating in the next iteration;
and step S103, stopping iteration when the iteration result meets the stop condition, and determining the processed fuzzy function obtained by the last iteration as the optimal fuzzy function.
In one implementation, the stop conditions include at least: the similarity of the processed clear images obtained by the two iterations is greater than a preset threshold value.
In one specific implementation, in the step S101, after the to-be-processed blurred image is received, a blur degree of the blurred fingerprint image may be determined. If the fuzzy degree of the fuzzy fingerprint image exceeds a preset critical value, continuing to execute the step S102; otherwise, directly carrying out fingerprint identification operation based on the fuzzy fingerprint image.
In one implementation, the step S101 may include: assigning a random value to each element in the matrix of the fuzzy function to obtain the initial fuzzy function.
For example, the random value may range from 0 to 1. That is, for each element in the matrix of the fuzzy function, a value is randomly selected from 0-1, so as to obtain the initial fuzzy function.
In a specific implementation, the step S101 may further include: determining the blurred image as the initial sharp image.
In one implementation, the blurred image may be imagery received by the sensor component.
In one implementation, the step S102 may include: a deconvolution operation is performed at each iteration, which may be based on the Maximum Likelihood (Maximum likehood) principle.
For example, at each iteration, the processed sharp image is calculated based on equation (2):
wherein k represents the kth iterative operation, and k is more than or equal to 0; f. ofk+1(x) A processed clear image is obtained for the (k + 1) th iteration; f. ofk(x) Processed sharp image obtained for the kth iterationAn image; g (x) is the blurred image; h isk(x) A processed fuzzy function obtained for the kth iteration; h isk(-x) is the inverse of the processed blur function obtained from the kth iteration; denotes a convolution operation; x represents a multiplication operation;
further, at each iteration, the processed fuzzy function is obtained based on formula (3):
wherein h isk+1(x) A processed fuzzy function obtained for the (k + 1) th iteration; f. ofkThe result of the inversion of the processed sharp image obtained in the kth iteration is (-x).
In one embodiment, the processed sharp image obtained in each iteration needs to satisfy the following constraints: the gray scale of each pixel in the processed clear image is larger than zero.
Further, the processed fuzzy function obtained by each iteration needs to satisfy the following constraint: the numerical value of each element in the processed fuzzy function is larger than zero; the processed fuzzy function meets the normalization condition; the number of rows and columns of the processed blur function remains unchanged.
The condition of meeting the normalization condition means that the sum of all elements in the matrix of the processed fuzzy function is 1.
In one implementation, after each iteration, a check operation may be performed to determine whether a stop condition has been reached. And if the similarity between the processed image obtained by the (k + 1) th iteration and the processed image obtained by the k-th iteration is greater than a preset threshold value, determining that the stop condition is met.
In particular, the preset threshold may be determined empirically, for example, as may be determined by previous experiments. Generally, when the iterative operation satisfies the stop condition, the average minimum number of errors is reached.
In one implementation, the similarity of the images may be characterized based on Structural Similarity Index (SSIM).
In one implementation, the stop condition may further include: the number of iterations reaches a preset number. So as to obtain a better fuzzy function by secondary check assistance.
For example, when the number of iterations reaches the preset number, but the similarity between the processed clear image obtained from the last iteration and the processed clear image obtained from the previous iteration is smaller than the preset threshold, the iterations are continued until the similarity between the processed clear images obtained from the previous iteration and the processed clear image obtained from the previous iteration is larger than the preset threshold.
For another example, when the number of iterations has not reached the preset number, if the similarity between the processed sharp image obtained from the latest iteration and the processed sharp image obtained from the previous iteration is greater than the preset threshold, the iteration may be immediately stopped.
Taking the clear fingerprint image (resolution 256 × 256) shown in fig. 2 as an example, the image actually captured by the sensor unit may be the blurred fingerprint image shown in fig. 3 due to the movement of the finger during imaging.
Theoretically, the blur function due to the horizontal movement of the finger is shown in fig. 4. The preferred blur function processed using the scheme of this embodiment is shown in fig. 5. It can be seen that with the scheme of the present embodiment, a preferred fuzzy function very close to the theoretical result can be obtained.
In a motion blur scenario, the resulting preferred blur function may be as shown in equation (4):
wherein rect () is a rectangular function; dsmearIs the moving distance of the collected object in a single exposure time.
In some embodiments, the linear motion of the captured object in a single exposure time may generate a smear in the image, and the larger the distance the captured object moves in the single exposure time, the more severe the smear phenomenon in the captured image. Specifically, the moving distance of the object to be captured within the single exposure time can be calculated according to the formula (5):
dsmear(t)＝υ·texp (5)
where upsilon is the moving speed of the collected object, texpIs the exposure time.
Therefore, the scheme of the embodiment can find out the fuzzy function caused by the movement of the object based on an iterative algorithm, so as to facilitate the later processing, such as deblurring. Specifically, a sharp image is estimated from a blurred image by deconvolution operation. Further, when the similarity of the clear images obtained by the current iteration and the later iteration is higher than a preset threshold value, the clear image obtained by the last iteration is good in restoration effect and tends to be stable, and the fuzzy function adopted when the deconvolution operation is performed on the iteration can be used as an optimal fuzzy function and used for deblurring processing of the currently acquired fuzzy image.
Compared with a scheme with a fixed iteration number as a stopping condition, the scheme with image similarity as the stopping condition is favorable for obtaining a better optimal fuzzy function. Specifically, since too many or too few iterations may affect the quality of the blur function and the degree of restoration of the blurred image, the deblurring effect of the image is not necessarily optimal for the optimal blur function obtained by processing with a fixed number of iterations as a stop condition. Based on the above, the scheme of the invention takes the image similarity as a stop condition, and determines the stop time according to the real-time processing effect in the iterative process, thereby obtaining a better optimal fuzzy function.
In one implementation, in response to having processed the preferred blur function, a deblurring operation may be performed on a blurred fingerprint image currently acquired by the sensor component based on the preferred blur function to obtain a corresponding sharp fingerprint image.
Specifically, a blurred image to be processed may be obtained, and then the blurred image to be processed is deblurred and restored to a corresponding sharp image based on the optimal blur function.
For example, the sharp image may be restored based on equation (6):
wherein,
to restore the resulting sharp image; g (x) is the blurred image to be processed; h (x) is the preferred blur function; *-1Representing a deconvolution operation.
In some embodiments, for the blurred fingerprint image acquired by the sensor component each time, the above method may be adopted to determine a preferred blur function suitable for the blurred fingerprint image acquired currently, and then perform a deblurring operation on the blurred fingerprint image acquired currently by the sensor component based on the preferred blur function to obtain a corresponding clear fingerprint image.
Fig. 6 shows a flowchart of a method for deblurring an image according to an embodiment of the present invention.
Step S201, acquiring a fuzzy image to be processed, and determining an initial fuzzy function and an initial clear image;
step S202, iteratively calculating based on a blurred image, a blur function and a sharp image to obtain a processed sharp image and a processed blur function, wherein the blur function in the first iteration is the initial blur function, the sharp image in the first iteration is the initial sharp image, and the processed sharp image and the processed blur function obtained in the previous iteration from the second iteration are used as the sharp image and the blur function for calculating in the next iteration;
and step S203, stopping iteration when the iteration result meets the stop condition, and determining a deblurring result based on the processed clear image obtained by the last iteration.
Wherein the stop condition includes at least: the similarity of the processed clear images obtained by the two iterations is greater than a preset threshold value.
For details of the iterative operation and the stopping condition, reference may be made to the related description in the embodiment shown in fig. 1, which is not repeated herein.
In one specific implementation, in the step S201, after the to-be-processed blurred image is received, a blur degree of the blurred fingerprint image may be determined. If the fuzzy degree of the fuzzy fingerprint image exceeds a preset critical value, continuing to execute the step S202; otherwise, directly carrying out fingerprint identification operation based on the fuzzy fingerprint image.
Specifically, the preset critical value in step S201 and step S101 of the foregoing embodiment may be determined according to whether the required information can be recognized from the original image. For example, if the probability that the fingerprint information cannot be identified from the blurred fingerprint image exceeds 80% when the blur degree of the blurred fingerprint image reaches a certain value, the value is determined as the preset critical value.
Taking the blurred fingerprint image (resolution 160 × 160) shown in fig. 7 as an example, the preferred blur function (size 9 × 9 pixels) shown in fig. 9 and the sharp fingerprint image shown in fig. 8 can be obtained simultaneously by using the above-mentioned embodiment shown in fig. 1. As can be seen from comparing fig. 7 and 8, the degree of blur of the fingerprint image is greatly reduced. In one embodiment, the number of iterations performed to obtain the clear fingerprint image shown in FIG. 8 is 10.
Fig. 10 is a schematic structural diagram of a blur function processing apparatus for an imaging system according to an embodiment of the present invention. Those skilled in the art will understand that the blur function processing apparatus 3 for an imaging system (hereinafter referred to as the blur function processing apparatus 3) described in the present embodiment can be used to implement the method described in the above embodiment shown in fig. 1.
Specifically, in this embodiment, the blur function processing apparatus 3 may include: an obtaining module 31, configured to obtain a blurred image, and determine an initial blur function and an initial sharp image; an iteration module 32, configured to iteratively perform an operation based on the blurred image, the blur function, and the sharp image to obtain a processed sharp image and a processed blur function, where the blur function in the first iteration is the initial blur function, the sharp image in the first iteration is the initial sharp image, and the processed sharp image and the processed blur function obtained in the previous iteration from the second iteration are used as the sharp image and the blur function for the operation in the next iteration; and the determining module 33 stops iteration when the iteration result meets the stop condition, and determines the processed fuzzy function obtained by the last iteration as the optimal fuzzy function.
In some embodiments, the stop condition includes at least: the similarity of the processed clear images obtained by the two iterations is greater than a preset threshold value.
For more details of the operation principle and the operation mode of the blur function processing apparatus 3, reference may be made to the related description in fig. 1, and details are not repeated here.
Fig. 11 is a schematic structural diagram of an image deblurring apparatus according to an embodiment of the present invention. Those skilled in the art will understand that the deblurring device 4 of the image (hereinafter referred to as the deblurring device 4) according to the present embodiment can be used to implement the method described in the above embodiment shown in fig. 6.
Specifically, the deblurring device 4 according to this embodiment may include: an obtaining module 41, configured to obtain a blurred image to be processed, and determine an initial blur function and an initial sharp image; a processing module 42, configured to iteratively perform an operation based on the blurred image, the blur function, and the sharp image to obtain a processed sharp image and a processed blur function, where the blur function in the first iteration is the initial blur function, the sharp image in the first iteration is the initial sharp image, and from the second iteration, the processed sharp image and the processed blur function obtained in the previous iteration are used as the sharp image and the blur function for the operation in the next iteration; and the determining module 43 stops iteration when the iteration result meets the stop condition, and determines the deblurring result based on the processed sharp image obtained by the last iteration.
In some embodiments, the stop condition includes at least: the similarity of the processed clear images obtained by the two iterations is greater than a preset threshold value.
For more details of the operation principle and the operation mode of the deblurring device 4, reference may be made to the related description in fig. 6, and details are not repeated here.
An embodiment of the present invention further provides an image capturing apparatus, including: the imaging system is used for acquiring a to-be-processed blurred image; and the deblurring module is coupled with the imaging system and is used for deblurring and restoring the blurred image to be processed into a corresponding clear image.
In some embodiments, the deblurring module is configured to perform the method shown in fig. 6 to deblur the blurred image to be processed into a corresponding sharp image.
In some embodiments, the image acquisition device may further comprise a blur function processing module coupled to the imaging system and the deblurring module, the blur function processing module being configured to perform the method of fig. 1 described above to determine the preferred blur function. And the optimal fuzzy function is transmitted to the deblurring module, and the deblurring module performs deblurring operation on the blurred image to be processed according to the optimal fuzzy function. For example, the deblurring module may obtain a corresponding sharp image based on equation (6) of the previous embodiment.
In a fingerprint identification scenario, the image capturing device may be an imaging device, such as a fingerprint imaging device, and the blurred image captured by the imaging system may be a blurred fingerprint image. Correspondingly, the deblurred and restored clear image obtained by the processing of the blur function processing module and the deblurring module is a clear fingerprint image.
By adopting the scheme of the embodiment, the acquired blurred image can be deblurred, and a clear image can be obtained without repeated acquisition. Therefore, the acquisition efficiency of the image acquisition equipment can be improved, and the acquisition time can be shortened.
Further, the embodiment of the present invention further discloses a storage medium, on which computer instructions are stored, and when the computer instructions are executed, the method technical solution described in the embodiment shown in fig. 1 or 6 is executed. Preferably, the storage medium may include a computer-readable storage medium such as a non-volatile (non-volatile) memory or a non-transitory (non-transient) memory. The storage medium may include ROM, RAM, magnetic or optical disks, etc.
Although the present invention is disclosed above, the present invention is not limited thereto. Various changes and modifications may be effected therein by one skilled in the art without departing from the spirit and scope of the invention as defined in the appended claims.
Claims (18)
1. A blur function processing method for an imaging system, comprising:
acquiring a blurred image, and determining an initial blurring function and an initial clear image;
iteratively calculating based on the blurred image, the blur function and the sharp image to obtain a processed sharp image and a processed blur function, wherein the blur function in the first iteration is the initial blur function, the sharp image in the first iteration is the initial sharp image, and the processed sharp image and the processed blur function obtained in the previous iteration from the second iteration are used as the sharp image and the blur function for calculating in the next iteration;
and when the iteration result meets the stop condition, stopping iteration, and determining the processed fuzzy function obtained by the last iteration as the optimal fuzzy function.
2. The blur function processing method according to claim 1, wherein the stop condition includes at least: the similarity of the processed clear images obtained by the two iterations is greater than a preset threshold value.
3. The blur function processing method of claim 1, wherein the blur function is represented in a matrix form, and the determining an initial blur function comprises:
assigning a random value to each element in the matrix to obtain the initial blur function.
4. The blur function processing method of claim 3, wherein the random value has a value in a range of 0 to 1.
5. The blur function processing method of claim 3 wherein the number of rows and columns of the matrix is determined according to the size of the light source of the imaging system.
6. The blur function processing method of claim 1, wherein the determining an initial sharp image comprises:
determining the blurred image as the initial sharp image.
7. The blur function processing method according to claim 1, wherein the iteratively operating based on the blurred image, the blur function, and the sharp image to obtain the processed sharp image and the processed blur function comprises:
and iteratively performing deconvolution operation based on the blurred image, the blur function and the sharp image to obtain a processed sharp image and a processed blur function.
8. The blur function processing method of claim 7, wherein the iteratively performing a deconvolution operation based on the blurred image, the blur function, and the sharp image to obtain a processed sharp image and a processed blur function comprises:
and calculating to obtain the processed clear image based on the following formula during each iteration:
wherein k represents the kth iterative operation, and k is more than or equal to 0; f. ofk+1(x) A processed clear image is obtained for the (k + 1) th iteration; f. ofk(x) A processed sharp image obtained for the kth iteration; g (x) is the blurred image; h isk(x) A processed fuzzy function obtained for the kth iteration; h isk(-x) is the inverse of the processed blur function obtained from the kth iteration; denotes a convolution operation; x represents a multiplication operation;
and obtaining the processed fuzzy function based on the following formula during each iteration:
wherein h isk+1(x) A processed fuzzy function obtained for the (k + 1) th iteration; f. ofkThe result of the inversion of the processed sharp image obtained in the kth iteration is (-x).
9. The blur function processing method according to claim 1, wherein the processed sharp image obtained in each iteration satisfies the following constraint:
the gray scale of each pixel in the processed clear image is larger than zero.
10. The blur function processing method of claim 1, wherein the processed blur function obtained from each iteration satisfies the following constraint:
the numerical value of each element in the processed fuzzy function is larger than zero;
the processed fuzzy function meets the normalization condition;
the number of rows and columns of the processed blur function remains unchanged.
11. The blur function processing method according to claim 1 or 2, wherein the stop condition includes: the number of iterations reaches a preset number.
12. The blur function processing method of claim 11, wherein stopping the iteration when the iteration result satisfies a stop condition comprises:
and when the iteration times reach the preset times and the similarity between the processed clear image obtained by the last iteration and the processed clear image obtained by the previous iteration is smaller than a preset threshold value, continuing the iteration until the similarity between the processed clear images obtained by the previous iteration and the processed clear image obtained by the next iteration is larger than the preset threshold value.
13. The blur function processing method according to any one of claims 1 to 12, further comprising:
acquiring a to-be-processed blurred image;
and deblurring and restoring the blurred image to be processed into a corresponding clear image based on the optimal blurring function.
14. The blur function processing method of claim 13, wherein the deblurring the blurred image to be processed into the corresponding sharp image based on the preferred blur function comprises: and restoring to obtain the clear image based on the following formula:
wherein,
to restore the resulting sharp image; g (x) is the blurred image to be processed; h (x) is the preferred blur function; *-1Representing a deconvolution operation.
15. A blur function processing apparatus for an imaging system, characterized by comprising:
the acquisition module is used for acquiring a blurred image and determining an initial blurring function and an initial sharp image;
the iteration module is used for performing iterative operation on the basis of the blurred image, the blur function and the clear image to obtain a processed clear image and a processed blur function, wherein the blur function in the first iteration is the initial blur function, the clear image in the first iteration is the initial clear image, and the processed clear image and the processed blur function obtained in the previous iteration from the second iteration are used as the clear image and the blur function for performing operation in the next iteration;
and the determining module stops iteration when the iteration result meets the stop condition and determines the processed fuzzy function obtained by the last iteration as the optimal fuzzy function.
16. The blur function processing apparatus of claim 15, wherein the stop condition comprises at least: the similarity of the processed clear images obtained by the two iterations is greater than a preset threshold value.
17. A storage medium having stored thereon computer instructions, wherein the computer instructions when executed perform the steps of the method of any one of claims 1 to 14.
18. An image acquisition apparatus, characterized by comprising:
the imaging system is used for acquiring a to-be-processed blurred image;
a blur function processing module coupled to the imaging system, the blur function processing module for performing the method of any of the above claims 1 to 14 to determine the preferred blur function;
and the deblurring module is coupled with the fuzzy function processing module and the imaging system and is used for deblurring and restoring the blurred image to be processed into a corresponding clear image based on the optimal fuzzy function.
Priority Applications (3)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010395146.6A CN113643192A (en) | 2020-05-11 | 2020-05-11 | Fuzzy function processing method and device for imaging system, image acquisition equipment and storage medium |
| TW110116805A TW202143101A (en) | 2020-05-11 | 2021-05-10 | Method and apparatus for processing image, imaging device and storage medium |
| US17/316,905 US20210350506A1 (en) | 2020-05-11 | 2021-05-11 | Method and apparatus for processing image, imaging device and storage medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010395146.6A CN113643192A (en) | 2020-05-11 | 2020-05-11 | Fuzzy function processing method and device for imaging system, image acquisition equipment and storage medium |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN113643192A true CN113643192A (en) | 2021-11-12 |
Family
ID=78415588
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202010395146.6A Pending CN113643192A (en) | 2020-05-11 | 2020-05-11 | Fuzzy function processing method and device for imaging system, image acquisition equipment and storage medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN113643192A (en) |
Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20070009169A1 (en) * | 2005-07-08 | 2007-01-11 | Bhattacharjya Anoop K | Constrained image deblurring for imaging devices with motion sensing |
| US20120155728A1 (en) * | 2010-12-20 | 2012-06-21 | General Electric Company | Tomographic iterative reconstruction |
| CN102542535A (en) * | 2011-11-18 | 2012-07-04 | 中国科学院自动化研究所 | Method for deblurring iris image |
| CN102708543A (en) * | 2012-04-19 | 2012-10-03 | 北京工商大学 | Fluorescent microscopic image restoring method based on blind deconvolution and sparse representation and device thereof |
| CN103744124A (en) * | 2013-12-31 | 2014-04-23 | 华中科技大学 | Infrared imaging detection locating method for underground tubular facility in flat terrain |
| CN107392882A (en) * | 2017-07-30 | 2017-11-24 | 湖南鸣腾智能科技有限公司 | A kind of method of the unzoned lens PSF iteration optimization initial values based on Corner Detection |
| CN108198151A (en) * | 2018-02-06 | 2018-06-22 | 东南大学 | A kind of star chart deblurring method based on improvement RL Deconvolution Algorithm Based on Frequency |
-
2020
- 2020-05-11 CN CN202010395146.6A patent/CN113643192A/en active Pending
Patent Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20070009169A1 (en) * | 2005-07-08 | 2007-01-11 | Bhattacharjya Anoop K | Constrained image deblurring for imaging devices with motion sensing |
| US20120155728A1 (en) * | 2010-12-20 | 2012-06-21 | General Electric Company | Tomographic iterative reconstruction |
| CN102542535A (en) * | 2011-11-18 | 2012-07-04 | 中国科学院自动化研究所 | Method for deblurring iris image |
| CN102708543A (en) * | 2012-04-19 | 2012-10-03 | 北京工商大学 | Fluorescent microscopic image restoring method based on blind deconvolution and sparse representation and device thereof |
| CN103744124A (en) * | 2013-12-31 | 2014-04-23 | 华中科技大学 | Infrared imaging detection locating method for underground tubular facility in flat terrain |
| CN107392882A (en) * | 2017-07-30 | 2017-11-24 | 湖南鸣腾智能科技有限公司 | A kind of method of the unzoned lens PSF iteration optimization initial values based on Corner Detection |
| CN108198151A (en) * | 2018-02-06 | 2018-06-22 | 东南大学 | A kind of star chart deblurring method based on improvement RL Deconvolution Algorithm Based on Frequency |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Abdelhamed et al. | A high-quality denoising dataset for smartphone cameras | |
| Hradiš et al. | Convolutional neural networks for direct text deblurring | |
| Jinno et al. | Multiple exposure fusion for high dynamic range image acquisition | |
| JP5651118B2 (en) | Image processing apparatus and image processing method | |
| Kee et al. | Modeling and removing spatially-varying optical blur | |
| EP2494524B1 (en) | Algorithms for estimating precise and relative object distances in a scene | |
| US8208746B2 (en) | Adaptive PSF estimation technique using a sharp preview and a blurred image | |
| Raghavendra et al. | Comparative evaluation of super-resolution techniques for multi-face recognition using light-field camera | |
| CN102907083A (en) | Image capturing apparatus, image processing apparatus, image processing method, and image processing program | |
| Quan et al. | Gaussian kernel mixture network for single image defocus deblurring | |
| KR102197089B1 (en) | Multi-task learning architecture for removing non-uniform blur of single image based on deep learning | |
| Duan et al. | Guided event filtering: Synergy between intensity images and neuromorphic events for high performance imaging | |
| Sada et al. | Image deblurring techniques—a detail review | |
| KR101528244B1 (en) | Processing of image data comprising effects of turbulence in a liquid medium | |
| CN107220945B (en) | Restoration method of multiple degraded extremely blurred image | |
| El-Henawy et al. | A comparative study on image deblurring techniques | |
| Xue | Blind image deblurring: a review | |
| Goilkar et al. | Implementation of blind and non-blind deconvolution for restoration of defocused image | |
| CN110009575B (en) | Infrared image stripe noise suppression method based on sparse representation | |
| CN113643192A (en) | Fuzzy function processing method and device for imaging system, image acquisition equipment and storage medium | |
| CN113643193A (en) | Image deblurring method and device, imaging equipment and storage medium | |
| Singh et al. | Comparative analysis of image deblurring techniques | |
| Bhagat et al. | Novel Approach to Estimate Motion Blur Kernel Parameters and Comparative Study of Restoration Techniques | |
| Sahay et al. | Shape extraction of low‐textured objects in video microscopy | |
| Li et al. | Joint motion deblurring with blurred/noisy image pair |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| WD01 | Invention patent application deemed withdrawn after publication | ||
| WD01 | Invention patent application deemed withdrawn after publication |
Application publication date: 20211112 |