WO2022141531A1

WO2022141531A1 - Image registration evaluation method and apparatus, and electronic device and readable storage medium

Info

Publication number: WO2022141531A1
Application number: PCT/CN2020/142430
Authority: WO
Inventors: 闫浩; 罗春; 李金升
Original assignee: 西安大医集团股份有限公司
Priority date: 2020-12-31
Filing date: 2020-12-31
Publication date: 2022-07-07
Also published as: CN116868233A

Abstract

Disclosed are an image registration evaluation method and apparatus, and an electronic device and a computer-readable storage medium. The image registration evaluation method comprises: acquiring a target registration position of a target image, wherein the target registration position refers to a position when the similarity between the target image and a preset reference image meets a preset registration condition during a registration process; and according to confidence level evaluation factors, determining a target confidence level of the target registration position, so as to evaluate the accuracy of image registration, wherein the confidence level evaluation factors are a plurality of factors for evaluating the accuracy of image registration. In the embodiments of the present application, a target confidence level of a target registration position can reflect the accuracy of the actual registration of a target image and a reference image, so that an image registration result with a low accuracy and even an incorrect image registration result can be prevented from being displayed, or a user can know the accuracy of an image registration result, thereby avoiding, to a certain extent, the problem of using an image registration result with a low accuracy and even an incorrect image registration result to perform further image processing.

Description

Image registration evaluation method, device, electronic device and readable storage medium

technical field

The present application relates to the field of medical technology, and in particular, to an image registration evaluation method, apparatus, electronic device, and computer-readable storage medium.

Background technique

Medical image registration technology is the basis of medical image processing, and plays a very important role in the fields of image information fusion, auxiliary diagnosis, surgical planning, and basic medical theory research.

Generally, image registration is the process of spatially matching two images. To register image A to image B, use B as a reference image and A as a floating image. By continuously moving floating image A and reference image B The registration is performed, and when the similarity between the floating image A and the reference image B is the largest, an image registration result is output, and the registration result may include the deformation field of the image A registered to the image B.

In the related art, regardless of the degree of accuracy of the image registration result, it is output to the user. However, the user does not know whether the image registration result is accurate enough, and will use the image registration result with low accuracy or even wrong (that is, the image registration result with very low accuracy) for further image processing, such as medical imaging deal with.

technical problem

As can be seen from the above, in the related art, the user cannot measure the accuracy of the image registration result, and it is easy to use the image registration result with low accuracy or even wrong for further image processing.

technical solutions

The present application provides an image registration evaluation method, device, electronic device and computer-readable storage medium, which saves time for image registration evaluation and improves the efficiency of image registration evaluation.

On the one hand, the present application provides an image registration evaluation method, the image registration evaluation method includes:

Obtaining a target registration position of the target image, wherein the target registration position refers to the position when the similarity between the target image and a preset reference image satisfies a preset registration condition during the registration process;

The target confidence of the target registration position is determined according to a confidence evaluation factor to evaluate the accuracy of image registration, wherein the confidence evaluation factor is a plurality of factors for evaluating the accuracy of image registration.

In some embodiments of the present application, the confidence evaluation factor includes at least one of an absolute value of the similarity measure, a change speed of the similarity measure, and a degree of fluctuation of the similarity measure.

In some embodiments of the present application, when the confidence evaluation factor includes the degree of fluctuation of the similarity measure, the determining the target confidence of the target registration position according to the confidence evaluation factor includes:

Obtain the similarity measure of the ith registration iteration between the target image and the reference image, wherein the target image is located at the target registration position during the ith registration iteration, and the i is greater than 0 the integer;

Obtain the similarity measure of the target image and the reference image for N registration iterations before the ith registration iteration and/or after the ith registration iteration, where N is greater than 0 the integer;

According to the similarity measure of the ith registration iteration and the similarity measure of the N registration iterations, determine the degree of fluctuation of N+1 similarity measures;

The target confidence level of the target registration position is determined according to the fluctuation degree of the N+1 similarity measures.

In some embodiments of the present application, the obtaining the similarity measure of the target image and the reference image for N registration iterations before the i-th registration iteration includes:

Obtain the similarity measure between the target image and the reference image from the i-Nth to the i-1th registration iteration, where N is less than or equal to the i.

In some embodiments of the present application, the acquiring the similarity measure of the target image and the reference image for N registration iterations after the ith registration iteration includes:

Obtain the similarity measure of the registration iteration from the i+1th to the i+Nth registration iteration between the target image and the reference image, where N is greater than the i.

In some embodiments of the present application, the acquiring the similarity measure of the target image and the reference image for N registration iterations before the ith registration iteration and after the ith registration iteration, include:

Obtain the similarity measure of the i-n1th to the i-1th registration iteration of the target image and the reference image, and the n1 is an integer less than or equal to the i;

Obtain the similarity measure of the registration iteration from the i+1th to the i+n2th time between the target image and the reference image, where the n2 is an integer greater than 0, and the sum of the n1 and the n2 is the N.

In some embodiments of the present application, the degree of fluctuation of the N+1 similarity measures is determined by any one of the following methods: the variance of the N+1 similarity measures, the standard deviation of the N+1 similarity measures, The range of the N+1 similarity measures, the interquartile difference of the N+1 similarity measures, the average difference of the N+1 similarity measures, or the N+1 similarity measures coefficient of variation.

In some embodiments of the present application, in the case that the confidence evaluation factor includes the change speed of the similarity measure, the determining the target confidence of the target registration position according to the confidence evaluation factor includes:

According to the similarity measure of the ith registration iteration and the similarity measure of the N registration iterations, determine the change speed of N+1 similarity measures;

The target confidence level of the target registration position is determined according to the change speed of the N+1 similarity measures.

In some embodiments of the present application, determining the change speed of N+1 similarity measures according to the similarity measure of the ith registration iteration and the similarity measure of the N registration iterations includes:

According to the similarity measure of the ith registration iteration and the similarity measure of the N registration iterations, constructing change curves of N+1 similarity measures;

obtaining each piecewise derivative of the variation curve;

The sum of the absolute values of the respective piecewise derivatives is obtained as the rate of change of the N+1 similarity measures.

In some embodiments of the present application, the determining the target confidence of the target registration position according to the confidence evaluation factor further includes:

Detecting whether the target confidence is greater than or equal to a preset reliability threshold;

When it is detected that the confidence of the target is greater than or equal to the preset confidence threshold, an image registration result is displayed, and the image registration result includes the target registration position of the target image and/or all registered positions of the target image. the target image and the reference image.

In some embodiments of the present application, it also includes:

When it is detected that the target confidence is less than the preset confidence threshold, prompt information indicating that the image registration accuracy of the target image and the reference image is low is output.

In some embodiments of the present application, the outputting prompt information indicating that the image registration accuracy of the target image and the reference image is low, and further includes:

Obtain the display instruction requesting to display the image registration result;

An image registration result of the target image and the reference image is displayed.

The image registration result of the target image and the reference image is displayed, and the target confidence is displayed.

On the other hand, the present application also provides an image registration evaluation device, the image registration evaluation device includes:

The acquiring unit is configured to acquire the target registration position of the target image, wherein the target registration position refers to the time when the similarity between the target image and the preset reference image satisfies the preset registration condition during the registration process. Location;

The evaluation unit is configured to determine the target confidence of the target registration position according to the confidence evaluation factor, so as to evaluate the accuracy of the image registration, wherein the confidence evaluation factor is a plurality of factors for evaluating the accuracy of the image registration. factor.

In some embodiments of the present application, the evaluation unit is specifically used for:

According to at least one of the absolute value of the similarity measure, the change speed of the similarity measure, and the fluctuation degree of the similarity measure, the target confidence level of the target registration position is determined to evaluate the accuracy of image registration.

obtaining the similarity measure of the registration iteration from the i-n1th to the i-1th registration iteration between the target image and the reference image, where n1 is an integer less than or equal to the i;

The degree of volatility of the N+1 similarity measures is determined by any of the following methods: variance of the N+1 similarity measures, standard deviation of the N+1 similarity measures, the N+1 similarity measures The range of the measure, the interquartile range of the N+1 similarity measures, the mean difference of the N+1 similarity measures, or the coefficient of variation of the N+1 similarity measures.

obtaining each piecewise derivative of the variation curve;

In some embodiments of the present application, the image registration evaluation apparatus further includes a display unit, and after the step of determining the target confidence of the target registration position according to the confidence evaluation factor, the display unit specifically uses At:

In some embodiments of the present application, the image registration evaluation apparatus further includes a prompt unit, and the display prompt is specifically used for:

In some embodiments of the present application, after the step of outputting prompt information indicating that the image registration accuracy of the target image and the reference image is low, the display unit is specifically configured to:

In some embodiments of the present application, after the step of determining the target confidence of the target registration position according to the confidence evaluation factor, the display unit is specifically configured to:

On the other hand, the present application also provides an electronic device, the electronic device comprising:

one or more processors;

memory; and

One or more application programs, wherein the one or more application programs are stored in the memory and configured to be executed by the processor to implement the steps in the image registration evaluation method described above.

On the other hand, the present application also provides a computer-readable storage medium on which a computer program is stored, and the computer program is loaded by a processor to execute the steps in the image registration evaluation method.

beneficial effect

In this application, the target confidence of the target registration position is determined according to the confidence evaluation factor to evaluate the accuracy of the image registration, because the target confidence of the target registration position can reflect the actual matching between the target image and the reference image to a certain extent. The accuracy of the image registration can be avoided to display the image registration results with low accuracy or even wrong, or it can let the user know the accuracy of the image registration results, which can avoid the use of low accuracy or even wrong registration results to a certain extent. The image registration results are subject to further image processing.

Description of drawings

In order to illustrate the technical solutions in the embodiments of the present application more clearly, the following briefly introduces the drawings that are used in the description of the embodiments. Obviously, the drawings in the following description are only some embodiments of the present application. For those skilled in the art, other drawings can also be obtained from these drawings without creative effort.

1 is a schematic diagram of a scene of an image registration evaluation system provided by an embodiment of the present application;

2 is a schematic flowchart of an embodiment of an image registration evaluation method provided by an embodiment of the present application;

3 is a schematic diagram of a scene of an image registration process provided in an embodiment of the present application;

FIG. 4 is a schematic diagram of the variation of the similarity measure of the ith registration iteration in the embodiment of the present application.

5 is a schematic structural diagram of an embodiment of an image registration evaluation device in an embodiment of the present application;

FIG. 6 is a schematic structural diagram of an embodiment of an electronic device in an embodiment of the present application.

Embodiments of the present application

The technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are only a part of the embodiments of the present application, but not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative work fall within the protection scope of the present application.

In the description of this application, it should be understood that the terms "center", "longitudinal", "lateral", "length", "width", "thickness", "upper", "lower", "front", " The orientation or positional relationship indicated by "rear", "left", "right", "vertical", "horizontal", "top", "bottom", "inside", "outside", etc. is based on the orientation shown in the drawings Or the positional relationship is only for the convenience of describing the present application and simplifying the description, rather than indicating or implying that the indicated device or element must have a specific orientation, be constructed and operated in a specific orientation, and therefore should not be construed as a limitation on the present application. In addition, the terms "first" and "second" are only used for descriptive purposes, and should not be construed as indicating or implying relative importance or implying the number of indicated technical features. Thus, features defined as "first", "second" may expressly or implicitly include one or more of said features. In the description of the present application, "plurality" means two or more, unless otherwise expressly and specifically defined.

In this application, the word "exemplary" is used to mean "serving as an example, illustration, or illustration." Any embodiment described in this application as "exemplary" is not necessarily to be construed as preferred or advantageous over other embodiments. The following description is presented to enable any person skilled in the art to make and use the present application. In the following description, details are set forth for the purpose of explanation. It is to be understood that one of ordinary skill in the art can realize that the present application may be practiced without the use of these specific details. In other instances, well-known structures and procedures have not been described in detail so as not to obscure the description of the present application with unnecessary detail. Therefore, this application is not intended to be limited to the embodiments shown but is to be accorded the widest scope consistent with the principles and features disclosed herein.

Embodiments of the present application provide an image registration evaluation method, apparatus, electronic device, and computer-readable storage medium, which will be described in detail below.

Please refer to FIG. 1. FIG. 1 is a schematic diagram of a scene of an image registration evaluation system provided by an embodiment of the present application. The image registration evaluation system may include an image acquisition device 100 and an electronic device 200, and a network of the image acquisition device 100 and the electronic device 200. Connected, the electronic device 200 is integrated with an image registration evaluation device, such as the electronic device in FIG.

In the embodiment of the present application, the electronic device 200 is mainly used to obtain the target registration position of the target image, where the target registration position refers to that the similarity between the target image and the preset reference image satisfies a predetermined level during the registration process. The position when the registration conditions are set; according to the confidence evaluation factor, the target confidence of the target registration position is determined to evaluate the accuracy of image registration, wherein the confidence evaluation factor is to evaluate the accuracy of image registration of multiple factors.

In the embodiment of the present application, the image acquisition device 100 may be a CT device, a CBCT device, or other medical imaging devices, such as an ultrasound device (such as a B-ultrasound device or a color ultrasound device), a magnetic resonance imaging (Magnetic Resonance Imaging, MRI) device, etc. Specifically, There is no limitation here. In a specific application scenario, the preset reference image is an image generated by CBCT in radiotherapy equipment, and the target image can be an image generated by CT equipment, CBCT equipment, or other medical imaging equipment (eg, ultrasound equipment, MRI equipment).

In this embodiment of the present application, the electronic device 200 may be a device including receiving and transmitting hardware, that is, a device having receiving and transmitting hardware capable of performing two-way communication on a two-way communication link. Such devices may include cellular or other communication devices with a single-line display or a multi-line display or a cellular or other communication device without a multi-line display. The specific electronic device 200 may specifically be a desktop terminal or a mobile terminal, and the electronic device 200 may specifically be one of a mobile phone, a tablet computer, a notebook computer, and the like.

Those skilled in the art can understand that the application environment shown in FIG. 1 is only an application scenario of the solution of the present application, and does not constitute a limitation on the application scenario of the solution of the present application. Other application environments may also include more than those shown in FIG. 1 . More or less electronic devices, or the network connection relationship of electronic devices, for example, only one electronic device and one image acquisition device are shown in FIG. other electronic devices, or/and one or more other image capturing devices connected to the electronic device network, specifically, for example, one electronic device 200 is connected to multiple image capturing devices 100, that is, one electronic device 200 displays multiple image capturing devices 100, or one image acquisition device 100 is connected to multiple electronic devices 200, that is, the images collected by one image acquisition device 100 are output to multiple electronic devices 200 for display, which is not specifically limited here.

In addition, as shown in FIG. 1 , the image registration evaluation system may further include a memory 300 for storing data, such as medical image data, such as medical image data collected by the image acquisition device 100 .

It should be noted that the schematic diagram of the scene of the image registration evaluation system shown in FIG. 1 is only an example, and the image registration evaluation system and the scene described in the embodiments of the present application are for the purpose of illustrating the technical solutions of the embodiments of the present application more clearly. This does not constitute a limitation on the technical solutions provided by the embodiments of the present application. Those of ordinary skill in the art know that with the evolution of the image registration evaluation system and the emergence of new business scenarios, the technical solutions provided by the embodiments of the present application are not suitable for similar technologies. question, the same applies.

First, an embodiment of the present application provides an image registration evaluation method, including: acquiring a target registration position of a target image, wherein the target registration position refers to the target image and a preset in the registration process. The position when the similarity of the reference image satisfies the preset registration condition; according to the confidence evaluation factor, the target confidence of the target registration position is determined to evaluate the accuracy of image registration, wherein the confidence evaluation factor Multiple factors for evaluating image registration accuracy.

As shown in FIG. 2, it is a schematic flowchart of an embodiment of the image registration evaluation method in the embodiment of the present application. The image registration evaluation method includes steps 201-202, wherein:

Step 201: Acquire the target registration position of the target image.

For two images in a set of medical image datasets, the image registration process simply maps one image (moving image) to another image (reference image, fixed image) by finding a spatial transformation , so that the points corresponding to the same position in space in the two pictures are in a one-to-one correspondence.

In the embodiment of the present application, the target image is a floating image in the image registration process, and the preset reference image is the reference image in the image registration process. Correspondingly, the target registration position is the position when the similarity between the target image and the preset reference image satisfies the preset registration condition during the image registration process, and the target registration position of the target image can be the position when the preset registration condition is met. The absolute position of the target image, such as the actual position of the target image in the image coordinate system, and the target registration position can also be the relative position of the target image when the preset registration conditions are met, such as the position of the target image relative to the reference image.

The concepts of "similarity between the target image and the preset reference image" and "preset registration conditions" are explained below:

Regarding the similarity between the target image and the preset reference image, it is the degree of similarity between the reference area of the reference image and the target area of the target image during the image registration process, which can be measured by the similarity measure between the reference area and the target area The similarity between the target image and the preset reference image. The similarity measure for measuring the similarity between the target image and the preset reference image includes: mutual information, mean square error, Kappa statistics, and the like.

Here, the reference area can be part or all of the image area in the reference image, and the target area is the image area that is searched for each time in the image registration process and compared with the reference image. The target area is varied.

Regarding the preset registration condition, it is the maximum similarity between the target image and the reference image during the image registration process, that is, the maximum similarity between the target area in the target image and the reference area in the reference image.

In step 202, there are various ways to obtain the target registration position of the target image, which are exemplary, including:

(1) Obtain the target registration position of the target image during the image registration process.

In the image registration process, when the similarity between the target image and the reference image satisfies the preset registration condition, the target registration position of the target image is obtained.

(2) Obtain the target registration position of the target image from the preset database.

During the image registration process, when the similarity between the target image and the preset reference image satisfies the preset registration condition, the target registration position of the target image is saved in the preset database. Obtaining the target registration position of the target image includes: obtaining the target registration position of the target image from a preset database.

It should be noted that the foregoing manner of acquiring the target registration position of the target image is only an exemplary description, and the embodiments of the present application are not limited to the foregoing manner of enumeration.

Step 202: Determine the target confidence of the target registration position according to the confidence evaluation factor, so as to evaluate the accuracy of the image registration.

First, a confidence evaluation factor is obtained, wherein the confidence evaluation factor is a plurality of factors for evaluating the accuracy of image registration, which may include at least one of a similarity measure, a change speed of the similarity measure, a fluctuation degree of the similarity measure, etc. By.

In an embodiment of the present application, the confidence evaluation factors can be preset and stored in the database. When the accuracy of image registration needs to be evaluated, the confidence evaluation factors are obtained from the database, and according to the confidence obtained from the database The degree evaluation factor is used to determine the target confidence of the target registration position.

After the confidence rating factor is obtained, the target confidence of the target registration position can be determined according to the confidence evaluation factor, so as to evaluate the accuracy of image registration. Exemplarily, the higher the target confidence level, for example, the target confidence level is 99%, it indicates that the image registration accuracy of the target image and the reference image is higher, and conversely, the target confidence level is lower, for example, the target confidence level is 60%. , it indicates that the image registration accuracy of the target image and the reference image is low.

In practical applications, in order to improve the speed of image registration, the target image will be locally searched for the area with the greatest similarity to the reference area in the reference image. When the search area in the target image has the greatest similarity with the reference area, The target image is considered to be in the target registration position. However, since the region with the greatest similarity to the reference region in the reference image is not globally searched in the target image, the target registration position is only the locally optimal registration position, not the globally optimal registration position. In this way, it is not accurate to evaluate the accuracy of image registration by similarity.

In the embodiment of the present application, in order to objectively and comprehensively evaluate the accuracy of image registration and improve the referentiality of image registration results, the concept of confidence is introduced to comprehensively evaluate the accuracy of the target registration position of the target image. . By obtaining the target registration position of the target image, the target confidence of the target registration position is determined according to the confidence evaluation factor, so as to evaluate the accuracy of image registration. Since the target confidence of the target registration position can reflect the accuracy of the actual registration between the target image and the reference image to a certain extent, compared with directly evaluating the accuracy of image registration by the similarity between the target image and the preset reference image, The target confidence level objectively evaluates the actual accuracy of the registration of the target image and the reference image from different angles. It allows users to know the accuracy of the image registration result, and to a certain extent avoids the problem of using the image registration result with low accuracy or even wrong for further image processing.

In the embodiment of the present application, the confidence evaluation factor can be implemented in multiple ways, for example, it may be one or more of the similarity measure, the change speed of the similarity measure, the fluctuation degree of the similarity measure, etc., which will be described with examples below.

(1) Confidence evaluation factors include similarity measure.

Since the similarity measure is an important factor reflecting the degree of image registration, it can reflect the accuracy of the image registration result to a certain extent. Therefore, the confidence evaluation factors including the similarity measure can evaluate the accuracy of image registration to a certain extent, and then allow users to understand the accuracy of image registration, and to a certain extent avoid the use of low accuracy or even wrong The image registration results are used for further image processing.

In some embodiments of the present application, the similarity between the target image and the reference image may be evaluated by a similarity measure. A similarity measure is a measure that comprehensively evaluates the degree of similarity between two things. In this embodiment of the present application, the similarity measure may have various forms, for example, mutual information, mean square error, kappa statistics, and the like.

In this embodiment of the present application, as shown in FIG. 3 , step 202 may specifically include steps 2021A to 2022A:

Step 2021A: Obtain a target similarity measure between the target image and the reference image at the target registration position.

In this embodiment of the present application, acquiring the similarity measure between the target image and the reference image at the target registration position may include: calculating the similarity measure between the reference area and the target area at the target registration position, as The similarity measure of the target image and the reference image at the target registration position.

Here, the value range of the similarity measure may be 0-1, or may be 0-100, and of course, it may also be set as required.

Step 2022A: Determine the target confidence of the target registration position according to the target similarity measure.

Further, step 2022A includes: determining the target confidence level according to the target similarity measure, the preset relationship between the similarity measure and the confidence level.

The preset relationship between the similarity measure and the confidence level may be a function of the confidence level on the similarity measure. For example, y=f(x), where x is the similarity measure and y is the confidence level.

In the first case, when the similarity measure between the target image and the reference image is positively correlated with the similarity, that is, the smaller the similarity measure between the target image and the reference image, the smaller the similarity between the target image and the reference image, the smaller the similarity between the target image and the reference image, The greater the similarity measure with the reference image, the greater the similarity between the target image and the reference image; at this time, the target confidence and the similarity measure are positively correlated, that is, the smaller the similarity measure, the smaller the target confidence, and the greater the similarity measure. larger, the greater the target confidence.

When the similarity measure is a positive value, step 2022A may specifically include step a1:

Step a1: Determine the similarity measure as the target confidence of the target registration position.

Exemplarily, the preset relationship between the similarity measure and the confidence level may be a function y=k*x of the confidence level on the similarity measure, where k is a preset coefficient greater than 0, x is the similarity measure, and y is the confidence level. Spend.

As an embodiment, the similarity measure may be the similarity measure of the target image and the reference image at the target registration position. For example, when the value of k in the function y=k*x is 1, if the similarity measure between the target image and the reference image at the target registration position is 1, the target confidence level of the target registration position is considered to be 1.

As another implementation manner, the similarity measure may also be the product of the similarity measure of the target image and the reference image and a preset coefficient at the target registration position. For example, when the value of k in the function y=k*x is non-1, such as the preset coefficient is 0.9, if the similarity measure between the target image and the reference image is 1 at the target registration position, then the target registration position The target confidence of is 1*0.9=0.9.

When the similarity measure is a negative value, step 2022A may specifically include steps b1-b2:

Step b1: Process the similarity measure to obtain a processed similarity measure.

Step b2: Determine the processed similarity measure as the target confidence of the target registration position.

Exemplarily, the preset relationship between the similarity measure and the confidence degree may be a function of the confidence degree on the similarity measure y=1-k*|x|, where k is a preset coefficient greater than 0, and x is greater than or equal to negative 1 , x is the similarity measure, and y is the confidence.

The processed similarity measure may be 1-k*|x|, where |x| is the absolute value of the similarity measure of the target image and the reference image at the target registration position.

As an embodiment, when the value of k in the function y=1-k*|x| is 1, (1-|x|) is used as the similarity measure after processing. Then, the processed similarity measure (1-|x|) is determined as the target confidence for the target registration location. For example, when the value of k in the function y=1-k*|x| is 1, if the similarity measure between the target image and the reference image at the target registration position is -0.2, it is considered that the target at the target registration position The confidence level is (1-|-0.2|)=0.8.

As another implementation manner, when the value of k in the function y=1-k*|x| is non-1, (1-k*|x|) is used as the similarity measure after processing. Then, the processed similarity measure (1-k*|x|) is determined as the target confidence for the target alignment location. For example, when the value of k in the function y=1-k*|x| is 0.9, if the similarity measure between the target image and the reference image at the target registration position is -0.2, it is considered that the target at the target registration position The confidence level is (1-0.9*|-0.2|)=0.82.

In the second case, when the similarity measure between the target image and the reference image is negatively correlated with the similarity, that is, the smaller the similarity measure between the target image and the reference image, the greater the similarity between the target image and the reference image, the greater the similarity between the target image and the reference image. The greater the similarity measure with the reference image, the smaller the similarity between the target image and the reference image; at this time, the target confidence and the similarity measure are negatively correlated, that is, the smaller the similarity measure, the greater the target confidence, and the greater the similarity measure. The larger the target confidence, the smaller the target confidence.

Exemplarily, when the similarity measure is mutual information, the similarity measure is usually a negative value, and the smaller the similarity measure is, the greater the similarity is; when the similarity measure is the mean square error, the similarity measure is usually a positive value, The smaller the similarity measure, the greater the similarity. It can be seen that the similarity measure can be a positive value or a negative value, but the smaller the similarity measure, the greater the similarity.

When the similarity measure is a negative value, step 2022A may specifically include steps c1-c2:

Step c1: Determine the similarity measure as the target confidence of the target registration position.

Exemplarily, the preset relationship between the similarity measure and the confidence level may be a function y=k*|x| of the confidence level on the similarity measure, where k is a preset coefficient greater than 0, x is the similarity measure, and y for confidence.

As an embodiment, the similarity measure may be the absolute value of the similarity measure of the target image and the reference image at the target registration position. For example, when the value of k in the function y=k*|x| is 1, if the similarity measure between the target image and the reference image at the target registration position is -1, the target confidence level of the target registration position is considered to be is 1.

As another implementation manner, the similarity measure may also be the product of the absolute value of the similarity measure of the target image and the reference image at the target registration position and a preset coefficient. For example, when the value of k in the function y=k*|x| is non-1, such as the preset coefficient is 0.9, if the similarity measure between the target image and the reference image at the target registration position is -1, then the target The target confidence for the registered position is 0.9*|-1|=0.9.

When the similarity measure is a positive value, step 2022A may specifically include steps d1-d2:

Step d1, processing the similarity measure to obtain a processed similarity measure.

Step d2, determining the processed similarity measure as the target confidence of the target registration position.

Exemplarily, the preset relationship between the similarity measure and the confidence degree may be a function of the confidence degree on the similarity measure y=1-k*x, where k is a preset coefficient greater than 0, x is less than or equal to 1, and x is Similarity measure, y is confidence.

The processed similarity measure may be y=1-k*x, where x is the absolute value of the similarity measure of the target image and the reference image at the target registration position.

As an embodiment, when the value of k in the function y=1-k*x is 1, (1-x) is used as the similarity measure after processing. Then, the processed similarity measure (1-x) is determined as the target confidence for the target registration location. For example, when the value of k in the function y=1-k*x is 1, if the similarity measure between the target image and the reference image at the target registration position is 0.2, the target confidence of the target registration position is considered to be (1-0.2)=0.8.

As another embodiment, when the value of k in the function y=1-k*x is non-1, (1-k*x) is used as the similarity measure after processing. Then, the processed similarity measure (1-k*x) is determined as the target confidence for the target registration location. For example, when the value of k in the function y=1-k*x is 0.9, if the similarity measure between the target image and the reference image at the target registration position is 0.2, then the target confidence of the target registration position is considered to be (1-0.9*0.2)=0.82.

(2) Confidence evaluation factors include the degree of fluctuation of the similarity measure.

Since the degree of fluctuation between the similarity measure of the target registration position and the similarity measure of other positions can reflect the degree of change between the similarity measure of the target registration position and the similarity measure of other positions to a certain extent, In turn, the registration accuracy of the target registration position can be reflected to a certain extent. Therefore, the confidence evaluation factors include the degree of fluctuation of the similarity measure, so that the user can know the accuracy of the image registration, and to a certain extent, avoid using the image registration results with low accuracy or even wrong for further image processing. .

Here, the similarity measure of the target registration position is the similarity measure of the target image and the reference image at the target registration position. The similarity measure at other locations is the similarity measure between the target image and the reference image at other locations. The other positions are the positions where the target image is located in addition to the target registration position during the image registration process.

When the confidence evaluation factor includes the degree of fluctuation of the similarity measure, step 202 may specifically include steps 2021B to 2024B:

Step 2021B: Obtain the similarity measure of the ith registration iteration between the target image and the reference image.

Among them, the target image is located at the target registration position during the ith registration iteration, and i is an integer greater than 0.

Referring to FIG. 3 , generally, the image registration process is essentially a process of continuously searching for similar regions in the target image. Each time a search is performed, that is, a position is moved, the similarity measure (between the target area and the reference area) is calculated once, and then the direction of the next move is determined according to this similarity measure and the similarity measures of all previous positions. This process is called a registration iteration.

Among them, the ith registration iteration is the process of searching and calculating the similarity measure for the ith in the process of image registration. The similarity measure of the ith registration iteration is the similarity measure of the target image and the reference image at the ith registration iteration.

In some embodiments, acquiring the similarity measure of the ith registration iteration between the target image and the reference image may specifically include: when the target image is located at the target registration position, calculating the similarity between the target area and the reference area metric as the similarity measure for the ith registration iteration.

Step 2022B: Obtain the similarity measure of the target image and the reference image for N registration iterations before the ith registration iteration and/or after the ith registration iteration.

Wherein, N is an integer greater than 0. The similarity measure of the N registration iterations includes the similarity measure of the target image and the reference image at each registration and registration iteration in the N registration iterations.

Step 2022B includes the following three situations:

In the first case, the similarity measure of N registration iterations before the ith registration iteration between the target image and the reference image is obtained. There are many ways to obtain them. The following examples illustrate:

(1) In an embodiment, the acquiring the similarity measure of the target image and the reference image for N registration iterations before the i-th registration iteration may specifically include: acquiring the target image A similarity measure for any N iterations preceding the ith registration iteration with the reference image. For example, N=5, the target image is located at the target registration position in the 10th registration iteration, the similarity measure of the first registration iteration, the similarity measure of the third registration iteration, and the sixth registration iteration can be obtained. The iterative similarity measure, the similarity measure of the seventh registration iteration, the similarity measure of the eighth registration iteration, and the similarity measure of a total of 5 registration iterations.

(2) In an embodiment, the acquiring the similarity measure of the target image and the reference image for N registration iterations before the i-th registration iteration may also specifically include: acquiring the target A similarity measure of the ith to ith registration iterations between the image and the reference image. where N is greater than i. For example, N=5, the target image is located at the target registration position in the 10th registration iteration, the similarity measure of the 5th registration iteration, the similarity measure of the 6th registration iteration, and the 7th registration iteration can be obtained. The iterative similarity measure, the similarity measure of the 8th registration iteration, the similarity measure of the ninth registration iteration, and the similarity measure of a total of 5 registration iterations.

In the second case, obtaining the similarity measure of the target image and the reference image for N registration iterations after the i-th registration iteration.

Specifically, it may include: acquiring similarity metrics of the target image and the reference image from the ith+1th to the ith+Nth registration iteration. There are many ways to obtain the similarity measure between the target image and the reference image from the i+1th to the i+Nth registration iteration. The following examples illustrate:

(1) In one embodiment, acquiring the similarity measure of the registration iteration from the i+1th to the i+Nth registration iteration between the target image and the reference image may specifically include: Set the region to perform local search, that is, search for any N target regions in the preset region, and calculate the similarity measure between the target region and the reference region from the i+1th to the i+Nth registration iteration, as N The similarity measure of the second registration iteration; wherein, the preset area is an image area within a preset range in the target image, and the image area includes the target area searched in the ith registration iteration.

(2) In an embodiment, acquiring the similarity measure of the registration iteration from the i+1th to the i+Nth registration iteration between the target image and the reference image may specifically include: Set the region to perform a global search, that is, traverse the entire preset region, and calculate the similarity measure between the target region and the reference region from the i+1th to the i+Nth registration iteration, as the N registration iterations where the preset area is an image area within a preset range in the target image, and the image area includes the target area searched in the ith registration iteration.

In the third case, to obtain the similarity measure of the target image and the reference image for N registration iterations before the ith registration iteration and after the ith registration iteration, there are many ways to obtain them.

In some implementation manners, the acquiring the similarity measure of the target image and the reference image for N registration iterations before the i-th registration iteration and after the i-th registration iteration may specifically be: Including steps e1～e2:

Step e1: Obtain the similarity measure of the registration iteration from the i-n1th to the i-1th registration iteration between the target image and the reference image.

Among them, n1 is an integer less than or equal to i.

For example, N=5, n1=3, the target image is located at the target registration position in the 10th registration iteration, the similarity measure of the seventh registration iteration, the similarity measure of the eighth registration iteration, the The similarity measure of 9 registration iterations, for a total of 3 registration iterations.

Step e2: Obtain the similarity measure of the registration iteration from the i+1th to the i+n2th registration iteration between the target image and the reference image.

Wherein, n2 is an integer greater than 0, and the sum of n1 and n2 is N.

Similar to obtaining the similarity measure of N registration iterations after the ith registration iteration between the target image and the reference image, obtain the similarity measure of the target image and the reference image from the i+1th to i+n2th registration iterations There are also many ways, which are listed below.

In an implementation manner, acquiring the similarity metric of the registration iteration from the i+1th to the i+n2th registration iteration between the target image and the reference image may specifically include: performing on a preset area in the target image Local search, that is, search for any N target areas in the preset area, and calculate the similarity measure between the target area and the reference area from the i+1th to i+n2th registration iterations respectively, and obtain n2 registration iterations where the preset area is an image area within a preset range in the target image, and the image area includes the target area searched in the ith registration iteration.

In an implementation manner, acquiring the similarity metric of the registration iteration from the i+1th to the i+n2th registration iteration between the target image and the reference image may specifically include: performing on a preset area in the target image From the i+1th to i+n2th registration iterations, and calculate the similarity measure of the target area and the reference area from the i+1th to i+n2th registration iterations respectively, and obtain the n2 registration iterations. similarity measure; wherein, the preset area is an image area within a preset range in the target image, and the image area includes the target area searched for in the ith registration iteration.

In some implementation manners, the acquiring the similarity measure of the target image and the reference image for N registration iterations before the i-th registration iteration and after the i-th registration iteration may specifically be: Including steps f1~f2:

Step f1, obtain the similarity measure of any n1 iterations before the ith registration iteration between the target image and the reference image.

Among them, n1 is an integer less than or equal to i. Obtaining the similarity measure between the target image and the reference image for any n1 iterations before the ith registration iteration is similar to obtaining the similarity measure for any N iterations between the target image and the reference image before the ith registration iteration. Specifically, you can Referring to the above descriptions and examples, details are not repeated here.

Step f2: Obtain the similarity measure of the registration iteration from the i+1th to the i+n2th registration iteration between the target image and the reference image.

Wherein, n2 is an integer greater than 0, and the sum of n1 and n2 is N. Step b2 is similar to the above-mentioned step a2. For details, reference may be made to the description and examples of the above-mentioned step a2, which will not be repeated here.

Step 2023B: Determine the degree of fluctuation of N+1 similarity measures according to the similarity measure of the ith registration iteration and the similarity measure of the N registration iterations.

In the embodiment of the present application, there are many ways to determine the degree of fluctuation of N+1 similarity measures, for example, the variance of N+1 similarity measures, the standard deviation of N+1 similarity measures, N+1 Range of similarity measures, interquartile range of N+1 similarity measures, mean difference of N+1 similarity measures, or coefficient of variation of N+1 similarity measures as N+1 similarity The degree of volatility of the measure.

Taking the N+1 similarity measures obtained from the i-n1th to the i+n2th registration iteration between the target image and the reference image as an example, we will introduce how to calculate the variance, N+ Standard deviation of 1 similarity measure, range of N+1 similarity measures, interquartile range of N+1 similarity measures, mean difference of N+1 similarity measures, or N+1 similarity The coefficient of variation for the measure.

1. Variance: It is one of the most commonly used measures to measure the degree of data fluctuation. The variance (sample variance) is the average of the squared values of the difference between each sample value and the average of the overall sample values.

As shown in Figure 4, the similarity measures of the target image and the reference image from the 1st to 9th registration iterations are: 0.5, 0.6, 0.5, 0.7, 0.6, 0.9, 0.8, 0.1, 0, respectively.

For example, i=6, N=4, N+1=5, n1=2, n2=2. If the target image is located at the target registration position in the sixth registration iteration, the similarity measure of the sixth registration iteration, the similarity measure of the two registration iterations before the sixth registration iteration, and the The similarity measure of the 2 registration iterations after the 6 registration iterations, the 5 similarity measures obtained from the 4th to the 8th registration iteration are respectively: 0.7, 0.6, 0.9, 0.8, 0.1.

At this time, the average value of the five similarity measures is: (0.7+0.6+0.9+0.8+0.1)/5=0.62, and the variance of the five similarity measures can be calculated as: [(0.7-0.62) ² +(0.6 -0.62) ² +(0.9-0.62) ² +(0.8-0.62) ² +(0.1-0.62) ² ]/5=0.097

2. Standard deviation: It is also one of the most commonly used measures to measure the degree of data fluctuation. The standard deviation is obtained by taking the square root of the variance.

At this point, the standard deviation of the 5 similarity measures is:

3. Extreme difference: also called full distance, refers to the difference between the maximum value and the minimum value in the data set, which can reflect the discrete situation of the data set to a certain extent.

At this time, the range of the five similarity measures is: 0.9-0.1=0.8.

4. Average difference: It is one of the values that indicates the degree of difference between the values of each variable, which can reflect the degree of fluctuation of a set of data to a certain extent. The mean difference is the arithmetic mean of the absolute value of the deviation of all units in the population from its arithmetic mean.

At this time, the average difference of (N+1)=5 similarity measures is: 0.216.

5. Coefficient of variation: It is used to measure the degree of fluctuation of categorical data, and to measure the degree of representation of the mode to a set of data. The coefficient of variation is the ratio of the standard deviation of the raw data to the mean of the raw data.

At this point, the coefficients of variation of the five similarity measures are:

The above variance, standard deviation, range, quartile, mean difference or coefficient of variation indicate that the fluctuation degree of N+1 similarity measures has their own advantages, and can be selected according to the actual application, and of course other functions can be selected. A statistical measure that reflects the degree of volatility.

Step 2024B: Determine the target confidence level of the target registration position according to the fluctuation degree of the N+1 similarity measures.

Among them, the target confidence is positively correlated with the fluctuation degree of N+1 similarity measures.

Determining the target confidence level of the target registration position according to the fluctuation degree of the N+1 similarity measures may specifically include: according to a preset relationship between the fluctuation degree and the confidence degree, and N+1 similarity measures The degree of fluctuation of the target is determined to determine the target confidence of the target registration position.

In one embodiment, the preset relationship between the fluctuation degree and the confidence level may be represented by a preset relationship mapping table. Exemplarily, there is a relationship between the degree of fluctuation and the degree of confidence as shown in Table 1 below. If the degree of fluctuation of the N+1 similarity measures is below 0.005, the target confidence degree can be determined to be 0; If the degree of fluctuation is above 0.05, the confidence level of the target can be determined to be 1.

Table 1

波动程度degree of volatility	置信度Confidence
0.005以下Below 0.005	00
0.0150.015	0.30.3
0.020.02	0.60.6
......	......
0.05以上0.05 or more	11

In one embodiment, the preset relationship between the fluctuation degree and the confidence level may be represented by a preset functional relationship. Exemplarily, the following functional relationship exists between the degree of fluctuation and the confidence: y=f(x)=20*x. Among them, y represents the confidence, and x represents the degree of volatility. If the fluctuation degree of N+1 similarity measures is 0.01, the target confidence level can be determined to be 0.2.

In the embodiment of the present application, it is considered that the degree of change between the similarity measure of the target registration position and the similarity measure of other positions can reflect the registration accuracy of the target registration position to a certain extent. The degree of data fluctuation can reflect the degree of change in the data, and the target is determined by the degree of fluctuation between N+1 similarity measures (including the similarity measures of N other positions and the similarity measure of the target registration position). The confidence of the registration position is used to evaluate the accuracy of image registration at the target registration position, which can avoid relying on a single similarity measure as the measurement standard of image registration, thereby improving the accuracy of image registration results to a certain extent. Accuracy.

(3) Confidence evaluation factors include the change speed of similarity measure.

Since the degree of fluctuation of the value (that is, the intensity of the change) is proportional to the amplitude of the fluctuation and inversely proportional to the length of the amplitude change, the fluctuation amplitude and the time of the amplitude change are divided, which is the speed of change in the physical sense. , and the rate of change of the value is still proportional to the degree of fluctuation. Therefore, the speed of change between the similarity measure of the target registration position and the similarity measure of other positions can reflect the degree of change between the similarity measure of the target registration position and the similarity measure of other positions to a certain extent , which can reflect the registration accuracy of the target registration position to a certain extent.

When the confidence evaluation factor includes the change speed of the similarity measure, as shown in FIG. 8 , step 202 may specifically include steps 2021C to 2024C:

2021C. Obtain the similarity measure of the ith registration iteration between the target image and the reference image.

Wherein, in the ith registration iteration, the target image is located at the target registration position, and i is an integer greater than 0. Step 2021C is similar to the above-mentioned step 2021B. For details, reference may be made to the description and examples of the above-mentioned step and step 2021B, which will not be repeated here.

2022C. Obtain similarity metrics of the target image and the reference image for N registration iterations before the ith registration iteration and/or after the ith registration iteration.

Wherein, N is an integer greater than 0. Step 2022C is similar to the above step 2022B. For details, reference may be made to the description and examples of the above step 2022B, which will not be repeated here.

2023C. Determine, according to the similarity measure of the ith registration iteration and the similarity measure of the N registration iterations, the change rates of N+1 similarity measures.

In some embodiments, determining the change speed of N+1 similarity measures according to the similarity measure of the ith registration iteration and the similarity measure of the N registration iterations may specifically include: according to the The similarity measure of the i-th registration iteration and the similarity measure of the N registration iterations are constructed, and the change curves of N+1 similarity measures are constructed; each piecewise derivative of the change curve is obtained; The sum of the absolute values of the respective piecewise derivatives is taken as the rate of change of the N+1 similarity measures.

Since the degree of curve fluctuation is proportional to the amplitude of the fluctuation and inversely proportional to the duration of the amplitude change, the division of the fluctuation amplitude and the time of the amplitude change is the slope in a physical sense. Therefore, the slope of the curve formed between the N+1 similarity measures can be used to reflect the change speed of the N+1 similarity measures.

In the embodiment of the present application, in order to observe the change speed of the N+1 similarity measures, a plurality of observation points are extracted from the change curves of the N+1 similarity measures, and the interval between two adjacent observation points is taken as a point. segments; compute the slope of each segment to determine the rate of change of the N+1 similarity measures.

Among them, the piecewise derivative refers to the slope between two adjacent observation points in the change curve of N+1 similarity measures.

The following takes N+1 similarity measures obtained from the i-n1th to i+n2th registration iterations between the target image and the reference image, and the similarity measure of each registration iteration as an observation point, to introduce how to Calculate the rate of change of N+1 similarity measures.

Please continue to refer to FIG. 4 , for example, i=6, N=4, N+1=5, n1=2, n2=2. If the target image is at the target registration position in the sixth registration iteration, take the similarity measure of the sixth registration iteration, the similarity measure of the two registration iterations before the sixth registration iteration, and the sixth registration iteration The similarity measures of the 2 registration iterations after the first registration iteration, the 5 similarity measures of the 4th to 8th registration iterations are obtained: 0.7, 0.6, 0.9, 0.8, 0.1.

According to 5 similarity measures: 0.7, 0.6, 0.9, 0.8, 0.1, the change curve of the similarity measure of points P4 to P8 as shown in Figure 4 can be constructed. Among them, P4, P5, P6, P7, and P8 represent the similarity measures of the 4th, 5th, 6th, 7th, and 8th registration iterations, respectively.

At this time, the piecewise derivative is the slope between the point where the similarity measure of the xth registration iteration is located and the point where the similarity measure of the x+1th registration iteration is located in the change curve of the similarity measure. Among them, 4≤x≤7. That is, there is a segment between P4 and P5, between P5 and P6, and between P7 and P8.

Calculate the slope between points P4 and P5: (0.7-0.6)/1=0.1, the slope between points P5 and P6: (0.6-0.9)/1=-0.3, the slope between points P6 and P7: (0.9-0.8)/1=0.1, slope between points P7 and P8: (0.8-0.1)/1=0.7. And calculate the sum of the absolute value of each piecewise derivative: 0.1+|-0.3|+0.1+0.7=1.2, as the change speed of 5 similarity measures.

2024C. Determine a target confidence level of the target registration position according to the change speed of the N+1 similarity measures.

Among them, the target confidence is positively correlated with the rate of change of N+1 similarity measures.

In some embodiments, determining the target confidence of the target registration position according to the change speed of the N+1 similarity measures may specifically include: according to a preset relationship between the change speed and the confidence, and N +1 rate of change of similarity measure to determine target confidence for target registration location.

In one embodiment, the preset relationship between the change speed and the confidence level may be represented by a preset relationship mapping table. Exemplarily, there is a relationship between the change speed and the confidence level as shown in Table 2 below. If the change speed of the N+1 similarity measures is 1.1, the target confidence level can be determined to be 0.1; if the change of the N+1 similarity measures If the speed is 1.3, the target confidence can be determined to be 0.3.

Table 2

变化速度speed of change	置信度Confidence
1.11.1	0.10.1
1.21.2	0.20.2

1.31.3	0.30.3
......	......

In one embodiment, the preset relationship between the change speed and the confidence level may be represented by a preset functional relationship. Exemplarily, the following functional relationship exists between the rate of change and the confidence: y=f(x)=0.1*x. where y is the confidence level and x is the rate of change. If the rate of change of the N+1 similarity measures is 1.1, the target confidence level can be determined to be 0.11.

In the embodiment of the present application, it is considered that the degree of change between the similarity measure of the target registration position and the similarity measure of other positions can reflect the registration accuracy of the target registration position to a certain extent. The speed of data change can reflect the severity of the data change, and the target is determined according to the speed of change between N+1 similarity measures (including the similarity measures of N other positions and the similarity measure of the target registration position). The confidence of the registration position is used to evaluate the accuracy of image registration at the target registration position, which can avoid relying on a single similarity measure as the measurement standard of image registration, thereby improving the accuracy of image registration results to a certain extent. Accuracy.

Further, in order to improve the accuracy of image registration evaluation, the target confidence determined when the confidence evaluation factor is the similarity measure, the target confidence determined when the confidence evaluation factor is the fluctuation degree of the similarity measure, and The confidence evaluation factor is the addition of two or three of the target confidences determined when the similarity measure changes speed according to a certain weight ratio, as the target confidence of the final target registration position.

For example, when the confidence evaluation factor is the similarity measure, the target confidence is 0.9; when the confidence evaluation factor is the fluctuation degree of the similarity measure, the target confidence is 0.8; the confidence evaluation factor is the change speed of the similarity measure. The target confidence level was determined to be 0.8. According to the preset weighting formula h=0.4*h1+0.3*h2+0.3*h3, the target confidence level of the final target registration position can be determined as: 0.4*0.9+0.3*0.8+0.3*0.8=0.84.

Among them, h1, h2, h3 respectively represent the target confidence determined when the above confidence evaluation factor is the similarity measure, the confidence evaluation factor is the target confidence determined when the confidence evaluation factor is the fluctuation degree of the similarity measure, and the confidence evaluation factor is the similarity The target confidence is determined when the metric changes speed, and h represents the target confidence of the final target registration position. 0.4, 0.3, and 0.3 are the weight coefficients of h1, h2, and h3, respectively. The weight coefficients of h1, h2, and h3 can be adjusted according to the actual situation and needs. This is just an example, not a limitation.

In some embodiments of the present application, in order to improve the referability of the image registration result, when the similarity between the target image and the preset reference image satisfies the preset registration conditions, that is, when the target registration position is located, The image registration result is output only if the target confidence at the registration position in front of the target is high. If the target confidence of the target registration position is low, the image registration result is not output. In order to avoid users adopting lower accuracy or even wrong image registration results for further image processing.

That is, in some embodiments of the present application, after step 202, it may further include: detecting whether the target confidence is greater than or equal to a preset confidence threshold; when it is detected that the target confidence is greater than or equal to the preset confidence When the degree threshold is reached, the image registration result is displayed, and the image registration result includes the target registration position of the target image and/or the registered target image and the reference image.

Further, in order to facilitate the user to understand the situation of image registration, when it is detected that the target confidence is less than the preset confidence threshold, output a message indicating that the image registration accuracy of the target image and the reference image is low. prompt information.

For example, the target confidence level ranges from 0% to 100%. The higher the confidence level, the higher the accuracy of the image registration result. When the target confidence level is less than a preset confidence threshold such as 50%, it is considered to be registered. The amount of information in the image data is insufficient, and the image registration result is not displayed. When the target confidence is greater than or equal to 50%, the image registration result is displayed, for example, the target registration position of the target image and/or the registered target image and the reference image are displayed.

The preset reliability threshold may be set according to actual requirements, which is not limited here.

Further, if the target confidence of the target registration position is low, the image registration result is not automatically displayed, but the user can choose to display the image registration result independently. At this time, the step of "outputting prompt information indicating that the image registration accuracy of the target image and the reference image is low" further includes: acquiring a display instruction requesting to display the image registration result; displaying the target image Image registration result with the reference image.

In some embodiments of the present application, regardless of the image registration result, the image registration result of the target image and the reference image is displayed, and the target confidence level of the target registration position is displayed. So that the user can know the accuracy of the displayed image registration result, and determine whether to use the displayed image registration result for further image processing according to actual needs. That is, in some embodiments of the present application, after step 202, it may further include: displaying an image registration result between the target image and the reference image, and displaying the target confidence.

In order to better implement the image registration evaluation method in the embodiment of the present application, on the basis of the image registration evaluation method, an image registration evaluation device is also provided in the embodiment of the present application, as shown in FIG. Evaluation device 500 includes:

The obtaining unit 501 is configured to obtain a target registration position of a target image, wherein the target registration position refers to when the similarity between the target image and a preset reference image satisfies a preset registration condition during the registration process s position;

The evaluation unit 502 is configured to determine the target confidence of the target registration position according to the confidence evaluation factor, so as to evaluate the accuracy of the image registration, wherein the confidence evaluation factor is a number of factors for evaluating the accuracy of the image registration. a factor.

In some embodiments of the present application, the evaluation unit 502 is specifically configured to:

Obtain the similarity measure of the registration iteration from the i+1th to the i+n2th time between the target image and the reference image, where the n2 is an integer of 0, and the sum of the n1 and the n2 is the said N.

obtaining each piecewise derivative of the variation curve;

In some embodiments of the present application, the image registration evaluation apparatus 500 further includes a display unit (not shown in the figure), and in the step of determining the target confidence of the target registration position according to the confidence evaluation factor After that, the display unit is specifically used for:

In some embodiments of the present application, the image registration evaluation apparatus 500 further includes a prompt unit (not shown in the figure), and the display prompt is specifically used for:

During specific implementation, the above units can be implemented as independent entities, or can be arbitrarily combined to be implemented as the same or several entities. The specific implementation of the above units can refer to the previous method embodiments, which will not be repeated here.

Since the image registration evaluation device can perform the steps in the image registration evaluation method in any embodiment of the present application, it can achieve the beneficial effects that can be achieved by the image registration evaluation method in any embodiment of the present application. description, which is not repeated here.

In addition, in order to better implement the image registration evaluation method in the embodiment of the present application, on the basis of the image registration evaluation method, the embodiment of the present application further provides an electronic device that integrates any of the methods provided in the embodiment of the present application. A kind of image registration evaluation device, the electronic equipment comprises:

one or more processors;

memory;

and one or more application programs, wherein the one or more application programs are stored in the memory and configured to be executed by the processor as described in any of the image registration evaluation method embodiments described above The steps in the image registration evaluation method.

The embodiments of the present application further provide an electronic device that integrates any of the image registration evaluation methods provided by the embodiments of the present application. As shown in FIG. 6, it shows a schematic structural diagram of an electronic device involved in an embodiment of the present application, specifically:

The electronic device may include a processor 601 of one or more processing cores, a memory 602 of one or more computer-readable storage media, a power supply 603 and an input unit 604 and other components. Those skilled in the art can understand that the structure of the electronic device shown in FIG. 6 does not constitute a limitation on the electronic device, and may include more or less components than the one shown, or combine some components, or arrange different components. in:

The processor 601 is the control center of the electronic device, uses various interfaces and lines to connect various parts of the entire electronic device, runs or executes the software programs and/or modules stored in the memory 602, and invokes the software programs stored in the memory 602. Data, perform various functions of electronic equipment and process data, so as to conduct overall monitoring of electronic equipment. Optionally, the processor 601 may include one or more processing cores; preferably, the processor 601 may integrate an application processor and a modem processor, wherein the application processor mainly processes the operating system, user interface, and application programs, etc. , the modem processor mainly deals with wireless communication. It can be understood that, the above-mentioned modulation and demodulation processor may not be integrated into the processor 601.

The memory 602 can be used to store software programs and modules, and the processor 601 executes various functional applications and data processing by running the software programs and modules stored in the memory 602 . The memory 602 may mainly include a stored program area and a stored data area, wherein the stored program area may store an operating system, an application program (such as a sound playback function, an image playback function, etc.) required for at least one function, and the like; Data created by the use of electronic equipment, etc. Additionally, memory 602 may include high-speed random access memory, and may also include non-volatile memory, such as at least one magnetic disk storage device, flash memory device, or other volatile solid state storage device. Accordingly, memory 602 may also include a memory controller to provide processor 601 access to memory 602 .

The electronic device also includes a power supply 603 for supplying power to various components. Preferably, the power supply 603 can be logically connected to the processor 601 through a power management system, so as to manage charging, discharging, and power consumption management functions through the power management system. Power source 603 may also include one or more DC or AC power sources, recharging systems, power failure detection circuits, power converters or inverters, power status indicators, and any other components.

The electronic device may also include an input unit 604 that may be used to receive input numerical or character information and generate keyboard, mouse, joystick, optical or trackball signal input related to user settings and function control.

Although not shown, the electronic device may further include a display unit and the like, which will not be described here. Specifically in this embodiment, the processor 601 in the electronic device loads the executable files corresponding to the processes of one or more application programs into the memory 602 according to the following instructions, and the processor 601 executes the execution and stores the executable files in the memory 602 . The application program in the memory 602, thereby realizing various functions, as follows:

Those of ordinary skill in the art can understand that all or part of the steps in the various methods of the above-mentioned embodiments can be completed by instructions, or by instructions that control relevant hardware, and the instructions can be stored in a computer-readable storage medium, and loaded and executed by the processor.

To this end, an embodiment of the present application provides a computer-readable storage medium, and the storage medium may include: a read-only memory (ROM, Read Only Memory), a random access memory (RAM, Random Access Memory), a magnetic disk or an optical disk, etc. . A computer program is stored thereon, and the computer program is loaded by the processor to execute the steps in any of the image registration evaluation methods provided in the embodiments of the present application. For example, the computer program being loaded by the processor may perform the following steps:

In the above-mentioned embodiments, the description of each embodiment has its own emphasis. For parts that are not described in detail in a certain embodiment, reference may be made to the above detailed description of other embodiments, and details are not repeated here.

During specific implementation, the above units or structures can be implemented as independent entities, or can be arbitrarily combined to be implemented as the same or several entities. For the specific implementation of the above units or structures, reference may be made to the foregoing method embodiments. No longer.

For the specific implementation of the above operations, reference may be made to the foregoing embodiments, and details are not described herein again.

The image registration evaluation method, device, electronic device, and computer-readable storage medium provided by the embodiments of the present application have been described in detail above. The principles and implementations of the present application are described with specific examples. The description of the embodiment is only used to help understand the method of the present application and its core idea; meanwhile, for those skilled in the art, according to the idea of the present application, there will be changes in the specific embodiments and application scope. As stated, the contents of this specification should not be construed as limiting the application.

Claims

An image registration evaluation method, characterized in that the image registration evaluation method comprises:

Obtaining a target registration position of the target image, wherein the target registration position refers to the position when the similarity between the target image and a preset reference image satisfies a preset registration condition during the registration process;

The target confidence of the target registration position is determined according to a confidence evaluation factor to evaluate the accuracy of image registration, wherein the confidence evaluation factor is a plurality of factors for evaluating the accuracy of image registration.
The image registration evaluation method according to claim 1, wherein the confidence evaluation factor comprises at least one of a similarity measure, a change speed of the similarity measure, and a fluctuation degree of the similarity measure.
The image registration evaluation method according to claim 2, wherein when the confidence evaluation factor includes the degree of fluctuation of the similarity measure, the determination of the target registration position according to the confidence evaluation factor Target confidence, including:

Obtain the similarity measure of the ith registration iteration between the target image and the reference image, wherein the target image is located at the target registration position during the ith registration iteration, and the i is greater than 0 the integer;

Obtain the similarity measure of the target image and the reference image for N registration iterations before the ith registration iteration and/or after the ith registration iteration, where N is greater than 0 the integer;

According to the similarity measure of the ith registration iteration and the similarity measure of the N registration iterations, determine the degree of fluctuation of N+1 similarity measures;

The target confidence level of the target registration position is determined according to the fluctuation degree of the N+1 similarity measures.
The image registration evaluation method according to claim 3, wherein the obtaining the similarity measure of the target image and the reference image for N registration iterations before the i-th registration iteration comprises:

Obtain the similarity measure between the target image and the reference image from the i-Nth to the i-1th registration iteration, where N is less than or equal to the i.
The image registration evaluation method according to claim 3, wherein the obtaining the similarity measure of the target image and the reference image N registration iterations after the i-th registration iteration comprises:

Obtain the similarity measure of the registration iteration from the i+1th to the i+Nth registration iteration between the target image and the reference image, where N is greater than the i.
The image registration evaluation method according to claim 3, wherein the acquisition of the target image and the reference image N before the ith registration iteration and after the ith registration iteration Similarity measures for sub-registration iterations, including:

obtaining the similarity measure of the registration iteration from the i-n1th to the i-1th registration iteration between the target image and the reference image, where n1 is an integer less than or equal to the i;

Obtain the similarity measure of the registration iteration from the i+1th to the i+n2th time between the target image and the reference image, where the n2 is an integer greater than 0, and the sum of the n1 and the n2 is the N.
The image registration evaluation method according to claim 3, wherein the degree of fluctuation of the N+1 similarity measures is determined by any of the following methods:

The variance of the N+1 similarity measures, the standard deviation of the N+1 similarity measures, the range of the N+1 similarity measures, the quartile of the N+1 similarity measures The disparity, the mean difference of the N+1 similarity measures, or the coefficient of variation of the N+1 similarity measures.
The image registration evaluation method according to claim 2, wherein when the confidence evaluation factor includes a change speed of the similarity measure, the determination of the target registration position according to the confidence evaluation factor Target confidence, including:

Obtain the similarity measure of the ith registration iteration between the target image and the reference image, wherein the target image is located at the target registration position during the ith registration iteration, and the i is greater than 0 the integer;

obtaining the similarity measure of the target image and the reference image for N registration iterations before the ith registration iteration and/or after the ith registration iteration, where N is greater than 0 the integer;

According to the similarity measure of the ith registration iteration and the similarity measure of the N registration iterations, determine the change speed of N+1 similarity measures;

The target confidence level of the target registration position is determined according to the change speed of the N+1 similarity measures.
The image registration evaluation method according to claim 8, characterized in that, according to the similarity measure of the ith registration iteration and the similarity measure of the N registration iterations, N+1 are determined. The rate of change of similarity measures, including:

According to the similarity measure of the ith registration iteration and the similarity measure of the N registration iterations, constructing change curves of N+1 similarity measures;

obtaining each piecewise derivative of the variation curve;

The sum of the absolute values of the respective piecewise derivatives is obtained as the rate of change of the N+1 similarity measures.
The image registration evaluation method according to claim 1, wherein the determining the target confidence of the target registration position according to the confidence evaluation factor, further comprising:

Detecting whether the target confidence is greater than or equal to a preset reliability threshold;

When it is detected that the confidence of the target is greater than or equal to the preset confidence threshold, an image registration result is displayed, and the image registration result includes the target registration position of the target image and/or all registered positions of the target image. the target image and the reference image.
The image registration evaluation method according to claim 10, further comprising:

When it is detected that the target confidence is less than the preset confidence threshold, prompt information indicating that the image registration accuracy of the target image and the reference image is low is output.
The image registration evaluation method according to claim 11, wherein the outputting prompt information indicating that the image registration accuracy of the target image and the reference image is low, further comprising:

Obtain the display instruction requesting to display the image registration result;

An image registration result of the target image and the reference image is displayed.
The image registration evaluation method according to claim 1, wherein the determining the target confidence of the target registration position according to the confidence evaluation factor, further comprising:

The image registration result of the target image and the reference image is displayed, and the target confidence is displayed.
An image registration evaluation device, characterized in that, the image registration evaluation device comprises:

The acquiring unit is configured to acquire the target registration position of the target image, wherein the target registration position refers to the time when the similarity between the target image and the preset reference image satisfies the preset registration condition during the registration process. Location;

The evaluation unit is configured to determine the target confidence of the target registration position according to the confidence evaluation factor, so as to evaluate the accuracy of the image registration, wherein the confidence evaluation factor is a plurality of factors for evaluating the accuracy of the image registration. factor.
An electronic device, characterized in that it comprises a processor and a memory, wherein a computer program is stored in the memory, and the processor executes the computer program according to any one of claims 1 to 13 when the processor calls the computer program in the memory Image Registration Evaluation Methods.
A computer-readable storage medium, characterized in that a computer program is stored thereon, and the computer program is loaded by a processor to execute the steps in the image registration evaluation method according to any one of claims 1 to 13 .