WO2022087846A1

WO2022087846A1 - Image processing method and apparatus, device, and storage medium

Info

Publication number: WO2022087846A1
Application number: PCT/CN2020/124113
Authority: WO
Inventors: 顾磊
Original assignee: Oppo广东移动通信有限公司
Priority date: 2020-10-27
Filing date: 2020-10-27
Publication date: 2022-05-05
Also published as: CN115885314A

Abstract

Provided in the embodiments of the present application are an image processing method and apparatus, a device, and a storage medium. The method comprises: acquiring an image to be subjected to identification, the image comprising a target object; identifying at least one quadrangle in the image to obtain coordinate information of each quadrangle; for each quadrangle, calculating a confidence level of the quadrangle according to the coordinate information of the quadrangle and by means of a Gaussian operation, wherein the confidence level is used for representing an assessment result of the quadrangle serving as a target area, and the target area is used for representing the form of the target object in the image; and determining at least one target quadrangle according to the confidence level of each quadrangle. A target quadrangle representing the form of a target object can be accurately identified, thereby providing a foundation for subsequent correction of the form of the target object so as to achieve a better presentation effect.

Description

Image processing method, device, device and storage medium

technical field

The embodiments of the present application relate to the technical field of image processing, and more particularly, to an image processing method, apparatus, device, and storage medium.

Background technique

With the deepening of the degree of electronization, users often need to collect images for some document objects through electronic devices including mobile phones. However, in the captured image, the tilted view effect of the document-type objects in the image is caused by the tilt of the electronic device during shooting.

In this regard, in the prior art, the image is often corrected by identifying the target quadrilateral in the image and according to the coordinate information of the target quadrilateral.

However, in the case that the image contains multiple quadrilaterals, or the photographed picture is relatively blurred, the existing technology cannot accurately identify the target quadrilateral, resulting in a poor effect of correcting the captured image.

SUMMARY OF THE INVENTION

Embodiments of the present application provide an image processing method, apparatus, device, and storage medium.

In a first aspect, an image processing method is provided, including:

obtaining an image to be identified, the image including the target object;

Identify at least one quadrilateral in the image to obtain coordinate information of each quadrilateral;

For each quadrilateral, according to the coordinate information of the quadrilateral, the reliability of the quadrilateral is calculated through the operation of the Gaussian function. The reliability is used to represent the evaluation result of the quadrilateral as the target area, and the target area is used to represent the target object in the image. form;

Based on the confidence of each quad, at least one target quad is determined.

In a second aspect, an image processing apparatus is provided, including:

an image acquisition unit for acquiring an image to be identified, the image including a target object;

an image recognition unit, used for recognizing at least one quadrilateral in the image to obtain coordinate information of each quadrilateral;

The image processing unit is used to calculate the reliability of the quadrilateral according to the coordinate information of the quadrilateral through Gaussian operation for each quadrilateral. The reliability is used to characterize the evaluation result of the quadrilateral as the target area, and the target area is used to represent the target object form in the image;

The image processing unit is further configured to determine at least one target quadrilateral according to the reliability of each quadrilateral.

In a third aspect, an electronic device is provided, including: a processor and a memory, where the memory is used for storing a computer program, and the processor is used for calling and running the computer program stored in the memory to execute the first aspect or each of its implementations. method.

In a fourth aspect, a computer-readable storage medium is provided for storing a computer program, and the computer program causes a computer to execute the method in the first aspect or each of its implementations.

In a fifth aspect, there is provided a computer program product comprising computer program instructions that cause a computer to perform a method as in the first aspect or implementations thereof.

In a sixth aspect, there is provided a computer program, the computer program causing a computer to perform the method as in the first aspect or implementations thereof.

According to the technical solution of the first aspect, by identifying at least one quadrilateral in the image, the coordinate information of each quadrilateral is obtained, so as to realize the preliminary screening of the quadrilateral; and based on the coordinate information of each quadrilateral, through Gaussian operation, the quadrilateral is obtained as For the credibility of the target area, by setting the preset parameters in the Gaussian function, the credibility of the quadrilateral approaching the target can be higher; and then one or more quadrilaterals with the highest credibility are selected as the target quadrilateral. The embodiment of the present application can accurately identify the target quadrilateral that can characterize the shape of the target object, which provides a basis for the subsequent correction of the shape of the target object to obtain a better presentation effect.

Description of drawings

FIG. 1 is a schematic flowchart of an image correction scene 100 provided by an embodiment of the present application;

FIG. 2 is a schematic frame diagram of an electronic device 200 provided by an embodiment of the present application;

FIG. 3 is a schematic flowchart of an image processing method 300 according to an embodiment of the present application;

FIG. 4 is a schematic diagram of a candidate graph 400 provided by an embodiment of the present application;

FIG. 5 is a schematic flowchart of an image processing method 500 according to an embodiment of the present application;

FIG. 6 is a schematic diagram of a to-be-recognized quadrilateral 600 provided by an embodiment of the present application;

FIG. 7 is a schematic flowchart of an image processing method 700 provided by an embodiment of the present application;

FIG. 8 shows a schematic block diagram of an image processing apparatus according to an embodiment of the present application;

FIG. 9 shows a schematic block diagram of an image processing apparatus according to an embodiment of the present application;

FIG. 10 shows a schematic block diagram of an image processing apparatus according to an embodiment of the present application;

FIG. 11 is a schematic structural diagram of an electronic device provided by an embodiment of the present application.

Detailed ways

The technical solutions in the embodiments of the present application will be described below with reference to the accompanying drawings in the embodiments of the present application. Obviously, the described embodiments are part of the embodiments of the present application, not all of the embodiments. With regard to the embodiments in the present application, all other embodiments obtained by persons of ordinary skill in the art without creative efforts shall fall within the protection scope of the present application.

At present, users often replace scanners and other devices with terminal devices such as mobile phones, tablets, computers, and wearable devices to capture images of document objects (the same as the target objects below), which improves the convenience of obtaining document objects. It is understood that document class objects include, but are not limited to, documents, business cards, posters, bulletin boards, rewritable whiteboards or blackboards, and the like. Because it is difficult to control the relative shooting angle between the terminal device and the target object, the document-type object is often presented as an inclined irregular quadrilateral due to the perspective transformation in the captured image. image is corrected. For example, by obtaining the inclination degree of the target quadrilateral related to the document class object in the image, and correcting the image according to the inclination degree of the target quadrilateral. However, in the actual image acquisition process, if the image acquisition capability of the terminal device is poor, the surrounding light is weak, or there are many interfering objects around the target object, etc., the quadrilaterals in the image are difficult to identify, or it is difficult to identify the number of polygons. Determine the target quadrilateral from the quadrilaterals.

The embodiments of the present application are applied to the above scenarios, in order to make the corrected image have a better display effect and accurately obtain the target quadrilateral in the image, with reference to FIG. The image 101 includes a plurality of line segments, and by identifying these line segments, quadrilaterals in the image, such as

quadrilaterals

102 and 103, are obtained, and by analyzing the reliability of each quadrilateral as a target area, and then based on the reliability of each quadrilateral It is determined that the target quadrilateral is the quadrilateral 102, and it should be understood that the target area is the boundary of the shape of the target object in the image. The embodiments of the present application achieve the technical effect of correctly acquiring the target quadrilateral in the image.

Wherein, the number of target quadrilaterals may be one to multiple. For example, when collecting images of multiple target objects at the same time, a target quadrilateral corresponding to each target object should be identified; or, for the same target object, multiple target quadrilaterals with high reliability may be determined, as an example instead of Restrictively, a final target quadrilateral may be determined based on a plurality of target quadrilaterals through algorithm screening, or a target quadrilateral selected by the user from the plurality of target quadrilaterals may be received as a final target quadrilateral.

As an example, a transformation matrix is obtained according to the determined at least one target quadrilateral, and perspective transformation is performed on the image according to the transformation matrix to obtain the transformed image 110 shown in FIG. 1 . The regular quadrilateral 112 is the effect after the quadrilateral 102 is transformed, and the target object is displayed in the image 110 in front after the quadrilateral 112 is transformed, which has a better display effect.

The technical solutions of the embodiments of the present application can be applied to various electronic devices to implement at least one of verification, optimization, and testing of language algorithm models. The electronic device may be a terminal device, such as a mobile phone (Mobile Phone), a tablet computer (Pad), a computer, a virtual reality (Virtual Reality, VR) terminal device, an augmented reality (Augmented Reality, AR) terminal device, an industrial control ( terminal equipment in industrial control, terminal equipment in self driving, terminal equipment in remote medical, terminal equipment in smart city or terminal equipment in smart home terminal equipment, etc. The terminal device in this embodiment of the present application may also be a wearable device, and a wearable device may also be called a wearable smart device, which is a general term for intelligently designing everyday wearable devices and developing wearable devices by applying wearable technology, such as Glasses, gloves, watches, clothing and shoes, etc. A wearable device is a portable device that is worn directly on the body or integrated into the user's clothing or accessories. Terminal devices can be stationary or mobile.

Exemplarily, the electronic device in this embodiment of the present application may also be a server. When the electronic device is a server, it can receive an image collected by the terminal device, and determine a target quadrilateral in the image.

FIG. 2 is a schematic structural diagram of an electronic device 200 according to an embodiment of the present application. As shown in FIG. 2 , the electronic device 200 includes: an image acquisition unit 210 , an image recognition unit 220 and an image processing unit 230 , and the image processing unit 230 includes at least a reliability calculation subunit 231 .

Wherein, the image acquisition unit 210 is used for acquiring the image to be recognized, and the image should contain a document class object. Exemplarily, an image collected by an image collection device, or an image transmitted by other devices, or an image input by a user may be received, which is not limited in this embodiment of the present application.

The image recognition unit 220 receives the image to be recognized sent by the image acquisition unit 210, and recognizes at least one quadrilateral in the image. Combining with FIG. 1, the quadrilateral 102 and the quadrilateral 103 can be obtained. Sent to the image processing unit 230.

The image processing unit 230 receives at least one quadrilateral sent by the image recognition unit 220, and determines the reliability of each quadrilateral through the reliability calculation subunit 231. Taking the example shown in FIG. 1, the reliability of the quadrilateral 102 is higher than that of the quadrilateral. 103 credibility.

Further, according to the reliability of each quadrilateral, the image processing unit 230 determines a quadrilateral with the highest reliability as the target quadrilateral, or after determining the reliability ranking, selects n quadrilaterals from the highest reliability as the target quadrilateral. The target quad.

As an example, the image processing unit 230 performs subsequent operations according to the determined at least one target quadrilateral. For example, at least one target quadrilateral is sent to the image display unit (not shown in the figure) for display, and the user can determine the recognition effect of the quadrilateral recognition by the electronic device according to the displayed at least one target quadrilateral, or make the user through the human-computer interaction interface Selecting a target quadrilateral as the final target quadrilateral; or, determining a transformation matrix according to at least one target quadrilateral, and performing perspective transformation on the image according to the transformation matrix. Exemplarily, when the number of target quadrilaterals is two or more, the image The processing unit 230 selects a final target quadrilateral from the two or more target quadrilaterals, or obtains the final target quadrilateral by performing a weighted average on the two or more target quadrilaterals.

The present application will be specifically described below through several embodiments.

FIG. 3 is a schematic flowchart of an image processing method 300 according to an embodiment of the present application.

In order to ensure that a quadrilateral closest to the inclination degree of the target object is obtained, the embodiment of the present application confirms the reliability of the quadrilateral as the target area, and obtains the target quadrilateral according to the reliability.

As shown in Figure 3, the image processing method includes:

S301: Acquire an image to be recognized.

It should be understood that the electronic device may capture an image of the target object through an image capturing device to obtain an image to be recognized, or may receive an image input by a user, or an image sent by other devices.

Exemplarily, after acquiring the image to be recognized, in order to facilitate subsequent recognition operations, the image may be preprocessed.

For example, perform a color space conversion on the image. The images collected by the image acquisition device are mostly RGB images, that is, color images formed by changing the three color channels of red R (Red), green G (Green), and blue B (Blue) and superimposing each other. In order to facilitate subsequent image recognition operations and improve processing efficiency, RGB images can be converted into color space in advance, such as converting to grayscale images or HSI images. ) and the brightness I (Intensity) to characterize the image.

Optionally, the preprocessing of the image to be identified further includes: performing edge detection on the image by any algorithm, such as the Canny algorithm and the Holistically-Nested Edge Detection (HED) algorithm.

Optionally, the preprocessing of the image to be recognized further includes: scaling the image to a preset size.

It should be understood that this embodiment does not limit the execution order of any of the foregoing preprocessing processes.

S302: Identify at least one quadrilateral in the image to obtain coordinate information of each quadrilateral.

The coordinate information of the quadrilateral includes coordinates of four vertices of the quadrilateral.

Exemplarily, in the process of identifying a quadrilateral in an image, first identify multiple line segments in the image, combine every four line segments in the multiple line segments into a candidate graphic, obtain all possible candidate graphics, and then analyze each candidate graphic. Whether the figure is a quadrilateral to be identified.

FIG. 4 is a schematic diagram of an image to be recognized 400 according to an embodiment of the present application. Exemplarily, this embodiment determines whether the candidate graphic satisfies the first preset condition, and when the candidate graphic satisfies the first preset condition, the candidate graphic is the quadrilateral to be identified, and specifically includes the following possible implementations:

1. The first preset condition is that the two first included angles of the candidate graphics are both smaller than the first preset value, and the first included angle is the included angle of any two non-adjacent line segments among the four line segments.

It is determined whether the two first included angles of the candidate graphics are smaller than the first preset value, and if both of the two first included angles are smaller than the first preset value, the candidate graphics is the quadrilateral to be identified. It should be understood that the first included angle is the included angle of the two included angles of the two line segments that is less than 90 degrees.

As shown in FIG. 4 , assuming that the candidate graphic is a candidate graphic composed of line segments AB, BC, CD and AD, then determine whether the angle between the line segment AB and the line segment CD is less than the first preset value, and the angle between the line segment AD and the line segment BC is determined. Whether it is smaller than the first preset value, if the included angle between line segment AB and line segment CD, and the included angle between line segment AD and line segment BC are all smaller than the first preset value, the candidate graphic is the quadrilateral to be identified.

Optionally, the first preset value is 30 degrees.

2. The first preset condition is that the four second included angles of the candidate graphics are all greater than the second preset value, and the second included angle is the included angle of any two adjacent line segments among the four line segments. Optionally, the first preset condition may be that the four second included angles of the candidate graphics are all smaller than the difference between 180 degrees and the second preset value.

As shown in FIG. 4 , assuming that the candidate graph is a candidate graph composed of line segments AB, BC, CD and AD, it is determined whether ∠A, ∠B, ∠C and ∠D are all greater than the second preset value. Exemplarily, the second preset value is 60 degrees, and it is determined whether ∠A, ∠B, ∠C, and ∠D are all greater than 60 degrees, that is, whether ∠A, ∠B, ∠C, and ∠D are all less than 120 degrees.

When ∠A, ∠B, ∠C and ∠D are all greater than the second preset value, the candidate figure is the quadrilateral to be identified.

3. The first preset condition is that the area ratio of the candidate graphic is greater than the fourth preset value, and the area ratio is the ratio of the area of the candidate graphic to the area of the image.

Optionally, the fourth preset value is one sixth.

Exemplarily, the first preset condition may also be any two or a combination of the above three examples. For example, when the two first included angles of the candidate graphics are both smaller than the first preset value, and the four second included angles of the candidate graphics are all greater than the second preset value, the candidate graphics is the quadrilateral to be identified; or the candidate graphics When the two first included angles of the candidate graphics are less than the first preset value, and the four second included angles of the candidate graphics are all greater than the second preset value, and the area ratio of the candidate graphics is greater than the fourth preset value, the candidate graphics is the quadrilateral to be identified.

Optionally, for a candidate graph composed of four unclosed line segments, such as the candidate graph in the upper left corner of the quadrilateral ABCD in Figure 4, determine the distance between the endpoints of the two unclosed line segments (the endpoint V and the endpoint W in Figure 4) Whether it is less than the preset distance, when the distance between the endpoints of the two line segments is less than the preset distance, determine whether the candidate graphic is a quadrilateral to be identified according to any of the above-mentioned embodiments, otherwise it is determined that the candidate graphic is not a quadrilateral to be identified .

Exemplarily, after determining that the candidate figure is the quadrilateral to be identified, the coordinates of the four vertices of the quadrilateral, that is, the coordinates of the four points A, B, C, and D in FIG. 4 are determined.

The quadrilateral obtained after screening through any of the above examples is closer to the target quadrilateral to be obtained, which reduces the scope for subsequent determination of the target quadrilateral and improves the processing efficiency.

As an example, in order to avoid interference such as noise, before each four line segments of the plurality of line segments are combined to obtain at least one candidate graphic in the above embodiment, the identified line segments may be optimized.

Exemplarily, it is determined whether there are at least two line segments on the same straight line among the identified multiple line segments, and the at least two line segments on the same straight line are merged. Optionally, through the coordinate information of each line segment, it can be identified that at least two line segments on the same straight line have the following possible positional relationships: 1. There is partial overlap, 2. One line segment is covered by another line segment, 3. The line segments are separated and the distance is less than the fourth preset value, and the distance is the distance between the closest endpoints of the two line segments. In this embodiment of the present application, the line segments on the same straight line that satisfy the above three possible positional relationships are merged. For example, the overlapping parts of the partially overlapping line segments are merged, the line segments covered by other line segments are deleted, and the line segment distance is smaller than the third line segment. Line segments with four preset values can be merged by translating one line segment to connect with another line segment, or supplementing the position between the line segments.

S303: For each quadrilateral, according to the coordinate information of the quadrilateral, the reliability of the quadrilateral is obtained by calculating the Gaussian operation.

Among them, the credibility is used to characterize the evaluation result of the quadrilateral as the target area, and the target area is used to characterize the shape of the target object in the image.

In this step, an evaluation is performed based on whether the quadrilateral can be used as the target area, so as to obtain the credibility of the quadrilateral. Exemplarily, at least one feature of the quadrilateral can be determined according to the coordinate information of the quadrilateral, and based on the at least one feature of the quadrilateral, the score of the quadrilateral as the target area is determined by Gaussian operation, that is, the credibility of the quadrilateral is obtained.

Optionally, the feature of the quadrilateral includes at least one of an area ratio feature, a first vertical angle feature, a second vertical angle feature, an adjacent angle feature, or a middle position feature. Among them, the area ratio feature is used to characterize the ratio of the area of the quadrilateral to the image area; the first vertical angle feature is used to characterize the angle between the midline of the quadrilateral and the vertical line of the image; the second vertical angle feature is used to characterize the midline of the quadrilateral The included angle with the projected gravity vector in the image; the adjacent angle feature is used to characterize the angular relationship of the adjacent corners of the quadrilateral; the middle position feature is used to characterize the coordinates of the middle position of the quadrilateral.

It should be understood that the Gaussian function used in the Gaussian operation includes three preset parameters: weight, average and variance. The average value is used to represent the expected eigenvalue of the corresponding feature, and the closer the eigenvalue is to the average value, the higher the calculated reliability. For example, when the ratio of the quadrilateral area to the image area is 0.8, which is most likely the target area, the average value is preset to 0.8.

S304: Determine at least one target quadrilateral according to the reliability of each quadrilateral.

In this step, the at least one identified quadrilateral is screened according to the reliability of the quadrilateral to obtain at least one target quadrilateral.

Possible implementation manner 1. Sort at least one quadrilateral according to the reliability of each quadrilateral, and use the n quadrilaterals with the highest reliability as at least one target quadrilateral, where n≥1, it should be understood that n The credibility of the quads can be the same or different, and the credibility of n quads is higher than the credibility of other quads that are not selected as target quads.

Possible implementation manner 2: Determine whether the reliability of each quadrilateral in the at least one quadrilateral is greater than a preset threshold, and determine the quadrilateral whose reliability is greater than the preset threshold as the target quadrilateral.

In the embodiment of the present application, by identifying at least one quadrilateral in the image, the coordinate information of each quadrilateral is obtained, so as to realize the preliminary screening of the quadrilateral; and based on the coordinate information of each quadrilateral, through Gaussian operation, the quadrilateral is obtained as the target area By setting the preset parameters in the Gaussian function, the reliability of the quadrilateral that is close to the target quadrilateral can be higher; and then one or more quadrilaterals with the highest reliability are selected as the target quadrilateral. The embodiment of the present application can accurately identify the target quadrilateral that can represent the shape of the target object, which provides a basis for the subsequent correction of the shape of the target object to obtain a better presentation effect.

On the basis of the above-mentioned embodiment, in order to accurately evaluate the reliability of the quadrilateral as the target area, the embodiment of the present application may perform reliability evaluation for each feature of the quadrilateral, and obtain the reliability corresponding to each feature. The reliability of each feature is summed to obtain the reliability of the quadrilateral.

FIG. 5 is a schematic flowchart of an image processing method 500 according to an embodiment of the present application. Exemplarily, as shown in FIG. 5 , based on the area ratio feature of the quadrilateral, the first vertical angle feature, the adjacent angle feature, the intermediate position feature, and the preset parameters corresponding to each feature, the reliability corresponding to each feature is obtained respectively. g, the reliability g corresponding to each feature is summed to obtain the reliability c of the quadrilateral.

It should be noted that, for the sake of clarity, the preset parameter corresponding to the area ratio feature in this embodiment is called the first preset parameter, the preset parameter corresponding to the first vertical angle feature is called the second preset parameter, and the preset parameter corresponding to the first vertical angle feature is called the second preset parameter. The preset parameter corresponding to the adjacent angle feature is called the third preset parameter, and the preset parameter corresponding to the middle position feature is called the fourth preset parameter.

Exemplarily, calculate the reliability g_1 corresponding to the area scale feature: determine the area of the quadrilateral according to the coordinate information of the quadrilateral, and use the ratio of the area of the quadrilateral to the area of the image as the feature value of the area scale feature, and then calculate according to the Gaussian function. Obtain the confidence corresponding to the area ratio feature. FIG. 6 is a schematic diagram of an image to be recognized 600 according to an embodiment of the present application. Combined with what is shown in Figure 6, according to the coordinates of the four vertices of the quadrilateral, that is, the coordinates of A, B, C, D and the pixel value of the image, that is, the length h and width w of the image in Figure 6, through the formula

Calculate the eigenvalue x ₁ of the area ratio feature, where AC and BD are the vectors of the diagonal of the quadrilateral.

Further, according to the calculated x ₁ , through the corresponding Gaussian function

Obtain the reliability g ₁ corresponding to the area ratio feature, wherein α ₁ , μ ₁ and σ ₁ are the first preset parameters corresponding to the area ratio feature, and α ₁ is used to characterize the reliability g ₁ corresponding to the area ratio feature The weight of , μ ₁ is the average value used to characterize the expected value of the area ratio, and σ ₁ is the variance.

Optionally, α ₁ =1, μ ₁ =0.7, σ ₁ =0.2.

Exemplarily, calculate the credibility g ₂ corresponding to the first vertical angle feature: according to the coordinate information of the quadrilateral, determine the vectors of the two center lines of the quadrilateral, and determine the angle between the vector of each center line and the vertical line, respectively, and The smallest included angle among the two included angles is used as the eigenvalue of the first vertical angle feature. It should be noted that the midline of the quadrilateral is the opposite side of the quadrilateral, that is, the connecting line between the midpoints of two non-adjacent sidelines, the vertical line is the straight line perpendicular to the horizontal sideline of the image, and the horizontal sideline refers to the image when the image is placed in the positive direction. , the edge in the horizontal direction.

As shown in Figure 6, N is the midpoint of the line segment AB, M is the midpoint of the line segment CD, P is the midpoint of the line segment BC, Q is the midpoint of the line segment AD, and the two midlines of the quadrilateral ABCD are MN and PQ respectively, It should be understood that the vector MN can be obtained by

Obtained, the vector PQ can be obtained by

get, and based on the formula

Obtain the first vertical angle feature x ₂ , where x=0 represents the y-axis in the two-dimensional coordinate system where the image is located, that is, the vertical line perpendicular to the horizontal edge of the image,

represents ∠b in Figure 6,

Indicates ∠a in Figure 6.

Further, according to the calculated x ₂ , through the corresponding Gaussian function

Obtain the reliability g ₂ corresponding to the first vertical angle feature, wherein α ₂ , μ ₂ and σ ₂ are the second preset parameters corresponding to the first vertical angle feature, and α ₂ is used to represent the first vertical angle feature corresponding to The weight of the reliability g ₂ , μ ₂ is the average value used to characterize the expected value of the first vertical angle, and σ ₂ is the variance.

Exemplarily, calculate the reliability g ₃ corresponding to the adjacent angle features: according to the coordinate information of the quadrilateral, determine the angles of the four vertex angles of the quadrilateral, and as shown in FIG. 6 , suppose ∠A=θ ₁ , ∠B=θ ₂ , ∠C=θ ₃ , ∠D=θ ₄ , further, a difference operation is performed on the vertex angles of every two vectors of the quadrilateral to obtain four adjacent angle differences, and then the four adjacent angle differences are calculated. And, the eigenvalue x ₃ of the adjacent angle feature is obtained, that is, x ₃ =|θ ₁ -θ ₂ |+|θ ₂ -θ ₃ |+|θ ₃ -θ ₄ |+|θ ₄ -θ ₁ |.

Further, according to the calculated x ₃ , through the corresponding Gaussian function

Obtain the reliability g ₃ corresponding to the adjacent angle feature, wherein α ₃ , μ ₃ and σ ₃ are the third preset parameters corresponding to the adjacent angle feature, and α ₃ is used to represent the reliability corresponding to the vector angle feature The weight of g ₃ , μ ₃ is the average value used to characterize the expected value of adjacent angles, and σ ₃ is the variance.

Exemplarily, calculate the reliability g ₄ corresponding to the feature of the middle position: according to the coordinate information of the quadrilateral, determine the coordinates of the middle position of the quadrilateral, it should be noted that the coordinates of the middle position are the average value of the coordinates of the four vertices of the quadrilateral , combined with Figure 6, the eigenvalues of the middle position feature, that is, the coordinates of the middle position of the quadrilateral

Further, according to the calculated x ₄ , through the corresponding Gaussian function

Obtain the reliability g ₄ corresponding to the middle position feature, wherein α ₄ , μ ₄ and σ ₄ are the fourth preset parameters corresponding to the middle position feature, and α ₄ is used to represent the reliability g ₄ corresponding to the vector angle feature The weight of , μ ₄ is the average value used to characterize the expected value of the middle position, and σ ₄ is the variance.

By formula c=Σg _i (x _i ,α _i ,μ _i ,σ _i ), i=1, 2, 3, 4, the reliability of the above features is summed to obtain the reliability c of the quadrilateral. It should be understood that the features of the quadrilateral listed in this embodiment are only examples, and more or less features may be included in practical applications.

In a possible implementation manner, the calculation of the reliability in any of the foregoing embodiments may also be combined with the reliability of the second vertical angle feature.

Exemplarily, calculating the reliability g ₅ of the second vertical angle feature: acquiring the pose information when the electronic device captures the image, where the pose information includes position information and/or attitude information, for example, directly read through the information carried by the image pose information, and then calculate the projected gravity vector g _p according to the pose information R _c and the formula g _p =P(R _c ,K,g ₀ ), where K is the camera internal parameter, and g ₀ is the world coordinate system Gravity vector, P is the projection function; further, according to the formula

in,

and

respectively represent the angle between the two midline vectors of the quadrilateral and the projected gravity vector, and take the minimum value of the two angles as the eigenvalue of the second vertical angle feature.

Further, according to the calculated x ₅ , through the corresponding Gaussian function

Obtain the reliability g ₅ corresponding to the second vertical angle feature, wherein α ₅ , μ ₅ and σ ₅ are preset parameters corresponding to the second vertical angle feature, also referred to as fifth preset parameters, and α ₅ is used to represent The weight of the reliability g ₅ corresponding to the vector angle feature, μ ₅ is the average value used to represent the expected value of the middle position, and σ ₅ is the variance.

FIG. 7 is a schematic flowchart of an image processing method 700 according to an embodiment of the present application.

The shooting scene of the image can reflect the type and characteristics of the target object to a certain extent. Then, setting different parameters for the Gaussian function for different shooting scenes can improve the accuracy of quadrilateral recognition.

For example, in the art gallery scene, the distance of the user viewing the painting is relatively short, so the area ratio of the captured target object in the image is larger, and the painting is generally in a hanging state, so the captured target object is generally presented in the image as Trapezoid, according to which the preset parameters corresponding to the area ratio feature and the preset parameters corresponding to the first vertical angle feature are set.

Optionally, the shooting scene includes but is not limited to: any one of an art gallery, an office building, and a school.

Exemplarily, the positioning information when the image is captured is first, and the shooting scene of the image is determined according to the positioning information, for example, the shooting scene is predicted according to the positioning information and the preset map information, and then the shooting scene is determined according to the determined shooting scene and the preset shooting scene. The corresponding relationship with the parameter group, determine the parameter group corresponding to the predicted shooting scene, the parameter group includes the preset parameters corresponding to one or more features in any of the above-mentioned embodiments, and then calculate the quadrilateral based on the determined parameter group. reliability.

Exemplarily, in the process of image acquisition, the positioning information of the electronic device is obtained through any positioning technology, for example, the positioning information of the electronic device is obtained through a global positioning system (Global Positioning System, GPS).

In addition to determining the shooting scene based on the positioning information, the embodiment of the present application can determine the shooting scene according to the user's intention, for example, receive the shooting scene selected by the user through the human-computer interaction interface, and then determine the corresponding parameter group according to the shooting scene.

Optionally, a setting instruction input by the user may be received, where the setting instruction includes a parameter group identifier, and the corresponding parameter group is determined according to the parameter group identifier indicated by the setting instruction.

The method embodiments of the present application are described in detail above with reference to FIGS. 1 to 7 , and the apparatus embodiments of the present application are described in detail below with reference to FIGS. 8 to 10 . It should be understood that the apparatus embodiments and the method embodiments correspond to each other, and are similar to For the description, refer to the method embodiment.

FIG. 8 shows a schematic block diagram of an image processing apparatus according to an embodiment of the present application. As shown in Figure 8, the image processing device 10 includes:

an image acquisition unit 11, for acquiring an image to be identified, the image including a target object;

The image recognition unit 12 is used to recognize at least one quadrilateral in the image, and obtain coordinate information of each quadrilateral;

The image processing unit 13 is used for each quadrilateral, according to the coordinate information of the quadrilateral, through Gaussian operation, calculates the credibility of the quadrilateral, the credibility is used to characterize the quadrilateral as the evaluation result of the target area, and the target area is used to represent the target. the shape of the object in the image;

The image processing unit 13 is further configured to determine at least one target quadrilateral according to the reliability of each quadrilateral.

The image processing device 10 in the embodiment of the present application includes an image acquisition unit 11 and an image recognition unit 12, and by identifying at least one quadrilateral in the image, the coordinate information of each quadrilateral is obtained, and the preliminary screening of the quadrilateral is realized; and based on The coordinate information of each quadrilateral is obtained through Gaussian operation to obtain the credibility of the quadrilateral as the target area. By setting the preset parameters in the Gaussian function, the credibility of the quadrilateral approaching the target quadrilateral can be higher; One or more quadrilaterals with the highest reliability are used as target quadrilaterals. The embodiment of the present application can accurately identify the target quadrilateral that can represent the shape of the target object, which provides a basis for the subsequent correction of the shape of the target object to obtain a better presentation effect.

Optionally, the image processing unit 13 is specifically used for:

For each feature of the quadrilateral, the reliability of the feature is determined according to the coordinate information of the quadrilateral and the Gaussian function corresponding to the feature, and the quadrilateral includes at least one feature;

Sum the confidences of at least one feature to get the confidences of the quadrilateral.

Optionally, the image processing unit 13 is specifically used for:

For each feature of the quadrilateral, determine the eigenvalue of the feature according to the coordinate information of the quadrilateral;

According to the feature value and the Gaussian function, the reliability of the feature is obtained, and the Gaussian function includes preset parameters corresponding to the feature.

Optionally, the image processing unit 13 is specifically used for:

According to the coordinate information of the quadrilateral, determine the area of the quadrilateral;

The ratio of the area of the quadrilateral to the area of the image is taken as the eigenvalue of the feature.

Optionally, the image processing unit 13 is specifically used for:

According to the coordinate information of the quadrilateral, the vectors of the two midlines of the quadrilateral are obtained;

Determine the angle between the vector of each midline and the vertical line, which is perpendicular to the horizontal edge of the image;

Take the smallest of the two included angles as the eigenvalue of the feature.

Optionally, the image processing unit 13 is specifically used for:

According to the coordinate information of the quadrilateral, determine the four vertex angles of the quadrilateral;

Perform a difference operation on every two adjacent vertex angles of the quadrilateral to obtain the difference of four adjacent angles;

The sum of the differences of four adjacent angles is taken as the eigenvalue of the feature.

Optionally, the image processing unit 13 is specifically used for:

According to the coordinate information of the quadrilateral, determine the coordinates of the middle position of the quadrilateral, and use the coordinates of the middle position as the feature value of the feature; the coordinates of the middle position are the average of the coordinates of the four vertices of the quadrilateral, and the coordinate information includes the coordinates of the four vertices .

Optionally, the preset parameters include weight, mean and variance.

Optionally, the image processing unit 13 is specifically used for:

Obtain pose information, which is used to characterize the position and/or attitude of the electronic device when the image is collected;

According to the pose information and the coordinate information of the quadrilateral, the eigenvalues of the features are determined.

Optionally, the image processing unit 13 is specifically used for:

According to the pose information and the projection function, the projected gravity vector is determined, and the projected gravity vector is used to represent the projection of the gravity vector in the world coordinate system in the coordinate system where the image is located;

Determine the angle between the vector of each midline of the quadrilateral and the projected gravity vector;

Take the smallest of the two included angles as the eigenvalue of the feature.

The data processing apparatus provided in the foregoing embodiments may execute the technical solutions of the foregoing method embodiments, and the implementation principles and technical effects thereof are similar, and are not repeated here.

Based on the embodiment shown in FIG. 8 , FIG. 9 shows a schematic block diagram of an image processing apparatus according to an embodiment of the present application. As shown in Figure 9, the image processing device 10 further includes:

a position obtaining unit 14, used for obtaining positioning information, the positioning information is used to represent the position of the electronic device when the image is collected;

The parameter determination unit 15 is used to determine the shooting scene of the image according to the positioning information;

The parameter determination unit 15 is further configured to determine a corresponding parameter group according to the shooting scene, where the parameter group includes preset parameters corresponding to at least one feature.

Based on the embodiment shown in FIG. 8 or FIG. 9 , FIG. 10 shows a schematic block diagram of an image processing apparatus according to an embodiment of the present application. As shown in FIG. 10 , the image processing apparatus 10 further includes: a receiving unit 16;

The receiving unit 16 is configured to receive the shooting scene input by the user;

Optionally, the image recognition unit 12 is specifically used for:

Identify at least one candidate graphic in the image, and the candidate graphic consists of four line segments;

For each candidate graphic, when the candidate graphic meets the first preset condition, it is determined that the candidate graphic is a quadrilateral, and coordinate information of the quadrilateral is obtained.

Optionally, the first preset condition includes at least one of the following:

The two first included angles of the candidate graphics are both smaller than the first preset value, and the first included angle is the included angle of any two non-adjacent line segments in the four line segments;

or,

The four second included angles of the candidate graphics are all greater than the second preset value, and the second included angle is the included angle of any two adjacent line segments in the four line segments;

or,

The area ratio of the candidate graphic is greater than the third preset value, and the area ratio is the ratio of the area of the candidate graphic to the area of the image.

Optionally, the image recognition unit 12 is specifically used for:

Identify multiple line segments in the image;

Combining every four line segments of the multiple line segments to obtain at least one candidate figure.

Optionally, the image recognition unit 12 is specifically used for:

Determine at least two line segments in the same straight line among the multiple line segments;

For each of the at least two line segments, when the two line segments satisfy the second preset condition, the two line segments are merged.

Optionally, the second preset condition includes:

There is a partial overlap between the two line segments;

Or, either of the two line segments is covered by the other;

Alternatively, the distance between the two line segments is smaller than the fourth preset value.

Optionally, the image processing unit 13 is specifically used for:

Sort at least one quadrilateral according to the reliability of each quadrilateral, and use the n quadrilaterals with the highest reliability as at least one target quadrilateral, n≥1;

or,

For each quadrilateral in the at least one quadrilateral, when the reliability of the quadrilateral is greater than a preset threshold, the quadrilateral is determined as the target quadrilateral.

FIG. 11 is a schematic structural diagram of an electronic device provided by an embodiment of the present application. The electronic device shown in FIG. 11 includes a processor 1210, and the processor 1210 can call and run a computer program from a memory, so as to implement the method in this embodiment of the present application.

Optionally, as shown in FIG. 11 , the electronic device 1200 may further include a memory 1220 . The processor 1210 may call and run a computer program from the memory 1220 to implement the methods in the embodiments of the present application.

The memory 1220 may be a separate device independent of the processor 1210, or may be integrated in the processor 1210.

Optionally, as shown in FIG. 11 , the electronic device 1200 may further include a transceiver 1230, and the processor 1210 may control the transceiver 1230 to communicate with other devices, specifically, may send information or data to other devices, or receive other devices Information or data sent by the device.

Among them, the transceiver 1230 may include a transmitter and a receiver. The transceiver 1230 may further include antennas, and the number of the antennas may be one or more.

Optionally, the electronic device 1200 may implement corresponding processes in each method of the embodiments of the present application, which are not repeated here for brevity.

It should be understood that the processor in this embodiment of the present application may be an integrated circuit chip, which has a signal processing capability. In the implementation process, each step of the above method embodiments may be completed by a hardware integrated logic circuit in a processor or an instruction in the form of software. The above-mentioned processor can be a general-purpose processor, a digital signal processor (Digital Signal Processor, DSP), an application specific integrated circuit (Application Specific Integrated Circuit, ASIC), an off-the-shelf programmable gate array (Field Programmable Gate Array, FPGA) or other available Programming logic devices, discrete gate or transistor logic devices, discrete hardware components. The methods, steps, and logic block diagrams disclosed in the embodiments of this application can be implemented or executed. A general purpose processor may be a microprocessor or the processor may be any conventional processor or the like. The steps of the method disclosed in conjunction with the embodiments of the present application may be directly embodied as executed by a hardware decoding processor, or executed by a combination of hardware and software modules in the decoding processor. The software modules may be located in random access memory, flash memory, read-only memory, programmable read-only memory or electrically erasable programmable memory, registers and other storage media mature in the art. The storage medium is located in the memory, and the processor reads the information in the memory, and completes the steps of the above method in combination with its hardware.

It can be understood that the memory in this embodiment of the present application may be a volatile memory or a non-volatile memory, or may include both volatile and non-volatile memory. Wherein, the non-volatile memory may be a read-only memory (Read-Only Memory, ROM), a programmable read-only memory (Programmable ROM, PROM), an erasable programmable read-only memory (Erasable PROM, EPROM), an electrically programmable read-only memory (Erasable PROM, EPROM). Erase programmable read-only memory (Electrically EPROM, EEPROM) or flash memory. Volatile memory may be Random Access Memory (RAM), which acts as an external cache. By way of illustration and not limitation, many forms of RAM are available, such as Static RAM (SRAM), Dynamic RAM (DRAM), Synchronous DRAM, SDRAM), double data rate synchronous dynamic random access memory (Double Data Rate SDRAM, DDR SDRAM), enhanced synchronous dynamic random access memory (Enhanced SDRAM, ESDRAM), synchronous link dynamic random access memory (Synchlink DRAM, SLDRAM) ) and direct memory bus random access memory (Direct Rambus RAM, DR RAM). It should be noted that the memory of the systems and methods described herein is intended to include, but not be limited to, these and any other suitable types of memory.

It should be understood that the above-mentioned memory is an example but not a limitative description, for example, the memory in this embodiment of the present application may also be a static random access memory (static RAM, SRAM), a dynamic random access memory (dynamic RAM, DRAM), Synchronous dynamic random access memory (synchronous DRAM, SDRAM), double data rate synchronous dynamic random access memory (double data rate SDRAM, DDR SDRAM), enhanced synchronous dynamic random access memory (enhanced SDRAM, ESDRAM), synchronous connection Dynamic random access memory (synch link DRAM, SLDRAM) and direct memory bus random access memory (Direct Rambus RAM, DR RAM) and so on. That is, the memory in the embodiments of the present application is intended to include but not limited to these and any other suitable types of memory.

Embodiments of the present application further provide a computer-readable storage medium for storing a computer program.

Optionally, the computer-readable storage medium can be applied to the electronic device in the embodiments of the present application, and the computer program enables the computer to execute the corresponding processes implemented by the electronic device in the various methods of the embodiments of the present application. For brevity, here No longer.

Embodiments of the present application also provide a computer program product, including computer program instructions.

Optionally, the computer program product can be applied to the electronic device in the embodiments of the present application, and the computer program instructions cause the computer to execute the corresponding processes implemented by the electronic device in the various methods of the embodiments of the present application. Repeat.

The embodiments of the present application also provide a computer program.

Optionally, the computer program can be applied to the electronic device in the embodiments of the present application. When the computer program runs on the computer, the computer executes the corresponding processes implemented by the electronic device in the various methods of the embodiments of the present application. For the sake of brevity. , and will not be repeated here.

Those of ordinary skill in the art can realize that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Skilled artisans may implement the described functionality using different methods for each particular application, but such implementations should not be considered beyond the scope of this application.

Those skilled in the art can clearly understand that, for the convenience and brevity of description, for the specific working process of the above-described devices, devices and units, reference may be made to the corresponding processes in the foregoing method embodiments, which will not be repeated here.

In the several embodiments provided in this application, it should be understood that the disclosed apparatus, apparatus and method may be implemented in other manners. For example, the apparatus embodiments described above are only illustrative. For example, the division of the above units is only a logical function division. In actual implementation, there may be other division methods. For example, multiple units or components may be combined or may be Integration into another system, or some features can be ignored, or not implemented. On the other hand, the shown or discussed mutual coupling or direct coupling or communication connection may be through some interfaces, indirect coupling or communication connection of devices or units, and may be in electrical, mechanical or other forms.

The units described above as separate components may or may not be physically separated, and components shown as units may or may not be physical units, that is, may be located in one place, or may be distributed to multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution in this embodiment.

In addition, each functional unit in each embodiment of the present application may be integrated into one processing unit, or each unit may exist physically alone, or two or more units may be integrated into one unit.

If the above functions are implemented in the form of software functional units and sold or used as independent products, they may be stored in a computer-readable storage medium. For such understanding, the technical solution of the present application can be embodied in the form of a software product in essence, or the part that contributes to the prior art or the part of the technical solution. The computer software product is stored in a storage medium, including Several instructions are used to cause a computer device (which may be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the methods in the various embodiments of the present application. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (Read-Only Memory, ROM), random access memory (Random Access Memory, RAM), magnetic disk or optical disk and other media that can store program codes .

The above contents are only specific embodiments of the present application, but the protection scope of the present application is not limited thereto. Any person skilled in the art who is familiar with the technical scope disclosed in the present application can easily think of changes or replacements, which should cover within the scope of protection of this application. Therefore, the protection scope of the present application shall be subject to the protection scope of the claims.

Claims

An image processing method, comprising:

acquiring an image to be identified, the image comprising a target object;

Identifying at least one quadrilateral in the image to obtain coordinate information of each quadrilateral;

For each quadrilateral, according to the coordinate information of the quadrilateral, through Gaussian operation, the reliability of the quadrilateral is calculated, and the reliability is used to represent the evaluation result of the quadrilateral as the target area. for characterizing the shape of the target object in the image;

Based on the confidence of each quad, at least one target quad is determined.
The method according to claim 1, wherein, for each quadrilateral, according to the coordinate information of the quadrilateral, calculating the reliability of the quadrilateral through Gaussian operation, comprising:

For each feature of the quadrilateral, the reliability of the feature is determined according to the coordinate information of the quadrilateral and the Gaussian function corresponding to the feature, and the quadrilateral includes at least one feature;

The confidence levels of the at least one feature are summed to obtain the confidence levels of the quadrilateral.
The method according to claim 2, wherein, for each feature of the quadrilateral, the reliability of the feature is determined according to coordinate information of the quadrilateral and a Gaussian function corresponding to the feature, include:

According to the coordinate information of the quadrilateral, determine the eigenvalue of the feature;

The reliability of the feature is obtained according to the feature value and the Gaussian function, where the Gaussian function includes preset parameters corresponding to the feature.
The method according to claim 3, wherein if the feature includes an area ratio feature, the determining the feature value of the feature according to the coordinate information of the quadrilateral comprises:

determining the area of the quadrilateral according to the coordinate information of the quadrilateral;

The ratio of the area of the quadrilateral to the area of the image is used as the feature value of the feature.
The method according to claim 3, wherein if the feature includes a first vertical angle feature, the determining the feature value of the feature according to the coordinate information of the quadrilateral comprises:

According to the coordinate information of the quadrilateral, the vectors of the two midlines of the quadrilateral are obtained;

Determine the angle between the vector of each midline and the vertical line, the vertical line being perpendicular to the horizontal edge of the image;

The smallest of the two included angles is taken as the eigenvalue of the feature.
The method according to claim 3, wherein, if the feature includes adjacent angle features, the determining the feature value of the feature according to the coordinate information of the quadrilateral comprises:

According to the coordinate information of the quadrilateral, determine the four vertex angles of the quadrilateral;

A difference operation is performed on every two adjacent vertex angles of the quadrilateral to obtain four adjacent angle differences;

The sum of four adjacent angle differences is taken as the feature value of the feature.
The method according to claim 3, wherein if the feature includes an intermediate position feature, the determining the feature value of the feature according to the coordinate information of the quadrilateral comprises:

According to the coordinate information of the quadrilateral, determine the coordinates of the middle position of the quadrilateral, and use the coordinates of the middle position as the feature value of the feature; the coordinates of the middle position are the coordinates of the four vertices of the quadrilateral The average value of , and the coordinate information includes the coordinates of the four vertices.
The method according to any one of claims 3 to 7, wherein the preset parameters include weight, average value and variance.
The method according to claim 3, wherein if the feature includes a second vertical angle feature, the determining the feature value of the feature according to the coordinate information of the quadrilateral comprises:

Obtaining pose information, the pose information is used to characterize the position and/or attitude of the electronic device when the image is collected;

According to the pose information and the coordinate information of the quadrilateral, the feature value of the feature is determined.
The method according to claim 9, wherein the determining the feature value of the feature according to the pose information and the coordinate information of the quadrilateral comprises:

According to the pose information and the projection function, determine the projection gravity vector, and the projection gravity vector is used to represent the projection of the gravity vector in the world coordinate system in the coordinate system where the image is located;

According to the coordinate information of the quadrilateral, the vectors of the two midlines of the quadrilateral are obtained;

Determine the angle between the vector of each center line of the quadrilateral and the projected gravity vector respectively;

The smallest of the two included angles is taken as the eigenvalue of the feature.
The method according to any one of claims 3 to 7, wherein the method further comprises:

obtaining positioning information, where the positioning information is used to represent the position of the electronic device when the image is collected;

determining the shooting scene of the image according to the positioning information;

According to the shooting scene, a corresponding parameter group is determined, and the parameter group includes preset parameters corresponding to at least one feature.
The method according to any one of claims 3 to 7, wherein the method further comprises:

Receive the shooting scene input by the user;

According to the shooting scene, a corresponding parameter group is determined, and the parameter group includes preset parameters corresponding to at least one feature.
The method according to any one of claims 1 to 7, wherein the identifying at least one quadrilateral in the image to obtain coordinate information of each quadrilateral includes:

Identify at least one candidate graphic in the image, and the candidate graphic is composed of four line segments;

For each candidate graphic, when the candidate graphic satisfies the first preset condition, it is determined that the candidate graphic is the quadrilateral, and the coordinate information of the quadrilateral is obtained.
The method according to claim 13, wherein the first preset condition comprises at least one of the following:

The two first included angles of the candidate graphics are both smaller than the first preset value, and the first included angle is the included angle of any two non-adjacent line segments in the four line segments;

or,

The four second included angles of the candidate graphics are all greater than the second preset value, and the second included angle is the included angle of any two adjacent line segments in the four line segments;

or,

The area ratio of the candidate graphic is greater than a third preset value, and the area ratio is a ratio of the area of the candidate graphic to the area of the image.
The method according to claim 13, wherein the identifying to obtain at least one candidate figure in the image comprises:

Identifying a plurality of line segments in the image;

Combining every four line segments of the plurality of line segments to obtain the at least one candidate figure.
The method according to claim 15, wherein before obtaining the at least one candidate figure by combining every four line segments of the plurality of line segments, the method further comprises:

determining at least two line segments in the same straight line among the plurality of line segments;

For each of the at least two line segments, when the two line segments satisfy a second preset condition, the two line segments are merged.
The method according to claim 16, wherein the second preset condition comprises:

The two line segments partially overlap;

Or, any one of the two line segments is covered by another line segment;

Alternatively, the distance between the two line segments is less than a fourth preset value.
The method according to any one of claims 1 to 7, wherein the determining at least one target quadrilateral according to the reliability of each quadrilateral comprises:

The at least one quadrilateral is sorted according to the reliability of each quadrilateral, and the n quadrilaterals with the highest reliability are used as the at least one target quadrilateral, and n≥1;

or,

For each quadrilateral in the at least one quadrilateral, when the reliability of the quadrilateral is greater than a preset threshold, the quadrilateral is determined to be the target quadrilateral.
An image processing device, comprising:

an image acquisition unit for acquiring an image to be identified, the image including a target object;

an image recognition unit for recognizing at least one quadrilateral in the image to obtain coordinate information of each quadrilateral;

The image processing unit is used for, for each quadrilateral, according to the coordinate information of the quadrilateral, through Gaussian operation, calculate the credibility of the quadrilateral, and the credibility is used to characterize the quadrilateral as the evaluation result of the target area , the target area is used to represent the shape of the target object in the image;

The image processing unit is further configured to determine at least one target quadrilateral according to the reliability of each quadrilateral.
The device according to claim 19, wherein the image processing unit is specifically configured to:

For each feature of the quadrilateral, the reliability of the feature is determined according to the coordinate information of the quadrilateral and the Gaussian function corresponding to the feature, and the quadrilateral includes at least one feature;

The confidence levels of the at least one feature are summed to obtain the confidence levels of the quadrilateral.
The device according to claim 20, wherein the image processing unit is specifically configured to:

According to the coordinate information of the quadrilateral, determine the eigenvalue of the feature;

The reliability of the feature is obtained according to the feature value and the Gaussian function, where the Gaussian function includes preset parameters corresponding to the feature.
The device according to claim 21, wherein the image processing unit is specifically configured to:

determining the area of the quadrilateral according to the coordinate information of the quadrilateral;

The ratio of the area of the quadrilateral to the area of the image is used as the feature value of the feature.
The device according to claim 21, wherein the image processing unit is specifically configured to:

According to the coordinate information of the quadrilateral, the vectors of the two midlines of the quadrilateral are obtained;

Determine the angle between the vector of each midline and the vertical line, the vertical line being perpendicular to the horizontal edge of the image;

The smallest of the two included angles is taken as the eigenvalue of the feature.
The device according to claim 21, wherein the image processing unit is specifically configured to:

According to the coordinate information of the quadrilateral, determine the four vertex angles of the quadrilateral;

A difference operation is performed on every two adjacent vertex angles of the quadrilateral to obtain four adjacent angle differences;

The sum of four adjacent angle differences is taken as the feature value of the feature.
The device according to claim 21, wherein the image processing unit is specifically configured to:

According to the coordinate information of the quadrilateral, determine the coordinates of the middle position of the quadrilateral, and use the coordinates of the middle position as the feature value of the feature; the coordinates of the middle position are the coordinates of the four vertices of the quadrilateral The average value of , and the coordinate information includes the coordinates of the four vertices.
The method according to any one of claims 21 to 25, wherein the preset parameters include weight, average value and variance.
The device according to claim 21, wherein the image processing unit is specifically configured to:

Obtaining pose information, the pose information is used to characterize the position and/or attitude of the electronic device when the image is collected;

According to the pose information and the coordinate information of the quadrilateral, the feature value of the feature is determined.
The device according to claim 27, wherein the image processing unit is specifically configured to:

determining a projected gravity vector according to the pose information and the projection function, where the projected gravity vector is used to represent the projection of the gravity vector in the world coordinate system in the coordinate system where the image is located;

According to the coordinate information of the quadrilateral, the vectors of the two midlines of the quadrilateral are obtained;

Determine the angle between the vector of each center line of the quadrilateral and the projected gravity vector respectively;

The smallest of the two included angles is taken as the eigenvalue of the feature.
The device according to any one of claims 21 to 25, wherein the device further comprises:

a position acquisition unit for acquiring positioning information, the positioning information being used to characterize the position of the electronic device when the image is collected;

a parameter determination unit, configured to determine the shooting scene of the image according to the positioning information;

The parameter determination unit is further configured to determine a corresponding parameter group according to the shooting scene, where the parameter group includes preset parameters corresponding to at least one feature.
The device according to any one of claims 21 to 25, wherein the device further comprises: a receiving unit;

The receiving unit is configured to receive a shooting scene input by a user;

The parameter determination unit is further configured to determine a corresponding parameter group according to the shooting scene, where the parameter group includes preset parameters corresponding to at least one feature.
The device according to any one of claims 19 to 25, wherein the image recognition unit is specifically configured to:

Identify at least one candidate graphic in the image, and the candidate graphic is composed of four line segments;

For each candidate graphic, when the candidate graphic satisfies the first preset condition, it is determined that the candidate graphic is the quadrilateral, and the coordinate information of the quadrilateral is obtained.
The device according to claim 31, wherein the first preset condition comprises at least one of the following:

The two first included angles of the candidate graphics are both smaller than the first preset value, and the first included angle is the included angle of any two non-adjacent line segments in the four line segments;

or,

The four second included angles of the candidate graphics are all greater than the second preset value, and the second included angle is the included angle of any two adjacent line segments in the four line segments;

or,

The area ratio of the candidate graphic is greater than a third preset value, and the area ratio is a ratio of the area of the candidate graphic to the area of the image.
The device according to claim 31, wherein the image recognition unit is specifically used for:

Identifying a plurality of line segments in the image;

Combining every four line segments of the plurality of line segments to obtain the at least one candidate figure.
The device according to claim 33, wherein the image recognition unit is specifically used for:

determining at least two line segments in the same straight line among the plurality of line segments;

For each of the at least two line segments, when the two line segments satisfy a second preset condition, the two line segments are merged.
The device according to claim 34, wherein the second preset condition comprises:

The two line segments partially overlap;

Or, any one of the two line segments is covered by another line segment;

Alternatively, the distance between the two line segments is less than a fourth preset value.
The device according to any one of claims 19 to 25, wherein the image processing unit is specifically configured to:

Sort the at least one quadrilateral according to the reliability of each quadrilateral, and use the n quadrilaterals with the highest reliability as the at least one target quadrilateral, where n≥1;

or,

For each quadrilateral in the at least one quadrilateral, when the reliability of the quadrilateral is greater than a preset threshold, the quadrilateral is determined to be the target quadrilateral.
An electronic device, characterized in that it comprises: a processor and a memory, the memory is used to store a computer program, the processor is used to call and run the computer program stored in the memory, and execute any one of claims 1 to 19. one of the methods described.
A computer-readable storage medium, characterized by being used for storing a computer program, the computer program causing a computer to perform the method according to any one of claims 1 to 19.
A computer program product comprising computer program instructions, the computer program instructions causing a computer to perform the method of any one of claims 1 to 19.
A computer program, characterized in that the computer program causes a computer to perform the method according to any one of claims 1 to 19.