Image processing method and device, computer device and storage medium

Abstract

The application relates to an image processing method and device and a computer device. The method comprises the following steps: obtaining a historical object image set and a current object image set corresponding to a target object; performing object point matching based on the historical object image set and the current object image set to obtain historical object matching points and current object matching points; when the historical object matching points and the current feature matching points are the same object points, performing spatial pose transformation parameter calculation on the historical object matching points to obtain historical spatial pose transformation parameters, and performing three-dimensional space matching point calculation to obtain historical feature point three-dimensional coordinates; performing spatial pose transformation parameter calculation on the current feature matching points to obtain current spatial pose transformation parameters, and performing three-dimensional space matching point calculation to obtain current feature point three-dimensional coordinates; and calculating a difference value corresponding to the historical feature point three-dimensional coordinates and the current feature point three-dimensional coordinates to obtain three-dimensional offset information corresponding to the target object. The method can improve the accuracy of deformation monitoring.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of image processing, and in particular to an image processing method and device, computer equipment, a storage medium and a computer program product. BACKGROUND

[0002] With the development of urbanization, in order to ensure construction safety in engineering construction, deformation monitoring needs to be performed on a building or a surrounding environment of the building. For example, target object deformation monitoring can be used to predict abnormal deformation such as cracks and landslides of the target object. A traditional deformation monitoring method uses sensors to monitor deformation of a target object.

[0003] However, the traditional deformation monitoring method needs to be in direct contact with the target object and can only monitor local deformation of the target object. When a large building is monitored and the sampling frequency is high, the accuracy of the sensor monitoring is low, thereby causing the problem of low accuracy of deformation monitoring. SUMMARY

[0004] Therefore, it is necessary to provide an image processing method, device, computer equipment, computer readable storage medium and computer program product capable of improving the accuracy of deformation monitoring to solve the above technical problems.

[0005] In a first aspect, the present application provides an image processing method. The method comprises:

[0006] obtaining a historical object image set and a current object image set corresponding to a target object, the historical object image set being obtained by an image acquisition device at a historical time from different acquisition positions for image acquisition of the target object, and the current object image set being obtained by the image acquisition device at a current time from different acquisition positions for image acquisition of the target object;

[0007] performing object point matching based on the historical object image set and the current object image set to obtain historical object matching points corresponding to each historical object image in the historical object image set and current object matching points corresponding to each current object image in the current object image set;

[0008] performing same object point matching using the historical object matching points corresponding to each historical object image and the current object matching points corresponding to each current object image;

[0009] when the historical object matching points and the current feature matching points are same object points, performing spatial pose transformation parameter calculation using the historical object matching points corresponding to each historical object image to obtain historical spatial pose transformation parameters, and performing three-dimensional spatial matching point calculation using the historical pose transformation parameters and the historical object matching points corresponding to each historical object image to obtain historical feature point three-dimensional coordinates;

[0010] The current spatial pose transformation parameter is calculated using the current feature matching points corresponding to each current object image, and the current spatial pose transformation parameter is obtained. The current feature point three-dimensional coordinates are calculated using the current pose transformation parameter and the current feature matching points corresponding to each current object image.

[0011] The difference between the historical feature point three-dimensional coordinates and the current feature point three-dimensional coordinates is calculated to obtain the three-dimensional offset information corresponding to the target object.

[0012] In a second aspect, the present application also provides an image processing device. The device comprises:

[0013] The acquisition module is configured to acquire a historical object image set and a current object image set corresponding to a target object. The historical object image set is obtained by an image acquisition device at a historical time from different acquisition positions for image acquisition of the target object. The current object image set is obtained by the image acquisition device at a current time from different acquisition positions for image acquisition of the target object.

[0014] The matching module is configured to perform object point matching based on the historical object image set and the current object image set, respectively, to obtain historical object matching points corresponding to each historical object image in the historical object image set and current object matching points corresponding to each current object image in the current object image set.

[0015] The matching point identification module is configured to perform the same object point matching using the historical object matching points corresponding to each historical object image and the current object matching points corresponding to each current object image.

[0016] The historical coordinate transformation module is configured to, when the historical object matching points and the current feature matching points are the same object points, perform spatial pose transformation parameter calculation using the historical object matching points corresponding to each historical object image to obtain historical spatial pose transformation parameters, and perform three-dimensional spatial matching point calculation using the historical pose transformation parameters and the historical object matching points corresponding to each historical object image to obtain historical feature point three-dimensional coordinates.

[0017] The current coordinate transformation module is configured to perform spatial pose transformation parameter calculation using the current feature matching points corresponding to each current object image to obtain current spatial pose transformation parameters, and perform three-dimensional spatial matching point calculation using the current pose transformation parameters and the current feature matching points corresponding to each current object image to obtain current feature point three-dimensional coordinates.

[0018] The offset calculation module is configured to calculate the difference between the historical feature point three-dimensional coordinates and the current feature point three-dimensional coordinates to obtain the three-dimensional offset information corresponding to the target object.

[0019] In a third aspect, the present application also provides a computer device. The computer device comprises a memory and a processor, the memory stores a computer program, and the processor implements the following steps when executing the computer program:

[0020] obtaining a historical object image set and a current object image set corresponding to a target object, the historical object image set being obtained by an image acquisition device at a historical time from different acquisition positions on the target object, and the current object image set being obtained by the image acquisition device at a current time from different acquisition positions on the target object;

[0021] performing object point matching based on the historical object image set and the current object image set respectively to obtain historical object matching points corresponding to each historical object image in the historical object image set and current object matching points corresponding to each current object image in the current object image set;

[0022] performing same object point matching using the historical object matching points corresponding to each historical object image and the current object matching points corresponding to each current object image;

[0023] when the historical object matching points and the current feature matching points are same object points, performing spatial pose transformation parameter calculation using the historical object matching points corresponding to each historical object image to obtain historical spatial pose transformation parameters, and performing three-dimensional spatial matching point calculation using the historical pose transformation parameters and the historical object matching points corresponding to each historical object image to obtain historical feature point three-dimensional coordinates;

[0024] performing spatial pose transformation parameter calculation using the current feature matching points corresponding to each current object image to obtain current spatial pose transformation parameters, and performing three-dimensional spatial matching point calculation using the current pose transformation parameters and the current feature matching points corresponding to each current object image to obtain current feature point three-dimensional coordinates;

[0025] calculating a difference value corresponding to the historical feature point three-dimensional coordinates and the current feature point three-dimensional coordinates to obtain three-dimensional offset information corresponding to the target object.

[0026] In a fourth aspect, the present application also provides a computer readable storage medium. The computer readable storage medium stores a computer program, and the computer program is executed by a processor to implement the following steps:

[0027] obtaining a historical object image set and a current object image set corresponding to a target object, the historical object image set being obtained by an image acquisition device at a historical time from different acquisition positions on the target object, and the current object image set being obtained by the image acquisition device at a current time from different acquisition positions on the target object;

[0028] perform object point matching based on the historical object image set and the current object image set respectively, to obtain historical object matching points corresponding to each historical object image in the historical object image set and current object matching points corresponding to each current object image in the current object image set;

[0029] perform same object point matching using the historical object matching points corresponding to each historical object image and the current object matching points corresponding to each current object image;

[0030] perform spatial pose transformation parameter calculation using the historical object matching points corresponding to each historical object image when the historical object matching points and the current feature matching points are same object points, to obtain historical spatial pose transformation parameters, and perform three-dimensional spatial matching point calculation using the historical pose transformation parameters and the historical object matching points corresponding to each historical object image, to obtain historical feature point three-dimensional coordinates;

[0031] perform spatial pose transformation parameter calculation using the current feature matching points corresponding to each current object image, to obtain current spatial pose transformation parameters, and perform three-dimensional spatial matching point calculation using the current pose transformation parameters and the current feature matching points corresponding to each current object image, to obtain current feature point three-dimensional coordinates;

[0032] calculate a difference value corresponding to the historical feature point three-dimensional coordinates and the current feature point three-dimensional coordinates, to obtain three-dimensional offset information corresponding to the target object.

[0033] In a fifth aspect, the present application further provides a computer program product. The computer program product comprises a computer program which, when executed by a processor, implements the following steps:

[0034] obtain a historical object image set and a current object image set corresponding to a target object, the historical object image set being obtained by image acquisition equipment at a historical time from different acquisition positions for image acquisition of the target object, and the current object image set being obtained by the image acquisition equipment at a current time from different acquisition positions for image acquisition of the target object;

[0035] perform object point matching based on the historical object image set and the current object image set respectively, to obtain historical object matching points corresponding to each historical object image in the historical object image set and current object matching points corresponding to each current object image in the current object image set;

[0036] perform same object point matching using the historical object matching points corresponding to each historical object image and the current object matching points corresponding to each current object image;

[0037] When the historical object matching point and the current feature matching point are the same object point, the historical spatial pose transformation parameter is calculated using the historical object matching point corresponding to each historical object image, the historical feature point three-dimensional coordinates are calculated using the historical pose transformation parameter and the historical object matching point corresponding to each historical object image, and the current spatial pose transformation parameter is calculated using the current feature matching point corresponding to each current object image, and the current feature point three-dimensional coordinates are calculated using the current pose transformation parameter and the current feature matching point corresponding to each current object image.

[0038] The current spatial pose transformation parameter is calculated using the current feature matching point corresponding to each current object image, the current feature point three-dimensional coordinates are calculated using the current pose transformation parameter and the current feature matching point corresponding to each current object image, and the historical feature point three-dimensional coordinates and the current feature point three-dimensional coordinates corresponding to the difference value are calculated, and the three-dimensional offset information corresponding to the target object is obtained.

[0039] The current spatial pose transformation parameter is calculated using the current feature matching point corresponding to each current object image, the current feature point three-dimensional coordinates are calculated using the current pose transformation parameter and the current feature matching point corresponding to each current object image, and the historical feature point three-dimensional coordinates and the current feature point three-dimensional coordinates corresponding to the difference value are calculated, and the three-dimensional offset information corresponding to the target object is obtained.

[0040] The above image processing method, device, computer equipment, storage medium and computer program product, by respectively matching object points of the historical object image set and the current object image set, obtaining the historical object matching point corresponding to each historical object image in the historical object image set and the current object matching point corresponding to each current object image in the current object image set. When it is detected that the historical object matching point and the current feature matching point are the same object point, the corresponding spatial pose transformation parameter is calculated according to the historical object matching point and the current feature matching point respectively, and then the historical feature point three-dimensional coordinates corresponding to the historical object matching point and the current feature point three-dimensional coordinates corresponding to the current feature matching point are calculated through the spatial pose transformation parameter. The historical feature point three-dimensional coordinates and the current feature point three-dimensional coordinates corresponding to the difference value are calculated, and the three-dimensional offset information corresponding to the target object is obtained. The historical feature point three-dimensional coordinates and the current feature point three-dimensional coordinates corresponding to the difference value are calculated, and the three-dimensional offset information corresponding to the target object is obtained. The same object point corresponding to the historical object matching point and the current feature matching point is converted from the image coordinates to the three-dimensional coordinates, the accuracy of the historical feature point three-dimensional coordinates and the current feature point three-dimensional coordinates is improved, and the three-dimensional offset information corresponding to the target object calculated through the historical feature point three-dimensional coordinates and the current feature point three-dimensional coordinates is more accurate, thereby improving the accuracy of the deformation monitoring of the target object. BRIEF DESCRIPTION OF DRAWINGS

[0041] Figure 1 It is an application environment diagram of the image processing method in one embodiment;

[0042] Figure 2 It is a flowchart of the image processing method in one embodiment;

[0043] Figure 3 It is a flowchart of obtaining three-dimensional offset information in one embodiment;

[0044] Figure 4 It is a specific flowchart of obtaining three-dimensional offset information in one embodiment;

[0045] Figure 5 is a schematic diagram of triangulation in one embodiment;

[0046] Figure 6 is a structural block diagram of an image processing apparatus in one embodiment;

[0047] Figure 7 is an internal structural diagram of a computer device in one embodiment;

[0048] Figure 8 is an internal structural diagram of a computer device in one embodiment. DETAILED DESCRIPTION

[0049] In order to make the purposes, technical solutions and advantages of the present application clearer, the present application is further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application and not used to limit the present application.

[0050] The image processing method provided by the embodiments of the present application can be applied to, for example, Figure 1The application environment shown. Among them, the terminal 102 communicates with the server 104 through the network. The data storage system can store the data required by the server 104 to process. The data storage system can be integrated on the server 104, or placed on the cloud or other network servers. The image acquisition device 106 is placed on the left and right sides of the target object A respectively, and the image acquisition device 106 is placed on the left and right sides of the target object A respectively. The image acquisition device 106 sends the collected image to the terminal 102 for processing. The terminal 102 acquires the historical object image set and the current object image set corresponding to the target object, the historical object image set is obtained by the image acquisition device 106 at the historical time from different collection positions to the target object, and the current object image set is obtained by the image acquisition device 106 at the current time from different collection positions to the target object; The terminal 102 respectively performs object point matching based on the historical object image set and the current object image set, obtains the historical object matching points corresponding to each historical object image in the historical object image set and the current object matching points corresponding to each current object image in the current object image set; The terminal 102 uses each historical object matching point corresponding to each historical object image and each current object matching point corresponding to each current object image to perform the same object point matching; When the historical object matching point and the current feature matching point are the same object point, the terminal 102 uses each historical object matching point corresponding to each historical object image to calculate the spatial pose transformation parameter, obtains the historical spatial pose transformation parameter, and uses the historical pose transformation parameter and each historical object matching point corresponding to each historical object image to calculate the three-dimensional space matching point, obtains the historical feature point three-dimensional coordinate; The terminal 102 uses each current object matching point corresponding to each current object image to calculate the spatial pose transformation parameter, obtains the current spatial pose transformation parameter, and uses the current pose transformation parameter and each current object matching point corresponding to each current object image to calculate the three-dimensional space matching point, obtains the current feature point three-dimensional coordinate; The terminal 102 calculates the difference value corresponding to the historical feature point three-dimensional coordinate and the current feature point three-dimensional coordinate, and obtains the three-dimensional offset information corresponding to the target object. Among them, the terminal 102 can be, but not limited to, various personal computers, notebook computers, smart phones, tablet computers, servers 104 can be implemented by independent servers or multiple servers composed of server clusters.

[0051] In one embodiment, as shown in Figure 2 An image processing method is provided. The embodiment is exemplified by the method applied to the terminal. It can be understood that the method can also be applied to the server, and can also be applied to the system including the terminal and the server, and can be realized through the interaction of the terminal and the server. In the embodiment, the method includes the following steps:

[0052] At step 202, a historical object image set and a current object image set corresponding to a target object are acquired. The historical object image set is obtained by image acquisition equipment at a historical time from different acquisition positions for image acquisition of the target object. The current object image set is obtained by image acquisition equipment at a current time from different acquisition positions for image acquisition of the target object.

[0053] At step 204, object point matching is performed based on the historical object image set and the current object image set, respectively, to obtain historical object matching points corresponding to each historical object image in the historical object image set and current object matching points corresponding to each current object image in the current object image set.

[0054] The target object refers to an object that is monitored for deformation such as cracking, landslides, etc., and can be a building, a mountain, or the like. The historical object matching points refer to object matching points between the historical object images, representing the same object point on the target object. The current object matching points refer to object matching points between the current object images, representing the same object point on the target object.

[0055] Specifically, the terminal acquires the historical object image set and the current object image set corresponding to the target object sent by the image acquisition equipment. The terminal performs feature extraction on each historical object image in the historical object image set and each current object image in the current object image set, respectively, to obtain historical feature object points corresponding to each historical object image and current feature object points corresponding to each current object image. Then the terminal performs feature matching on the historical feature object points corresponding to each historical object image to obtain historical object matching points corresponding to each historical object image. The terminal performs feature matching on the current feature object points corresponding to each current object image to obtain current object matching points corresponding to each current object image.

[0056] At step 206, the same object point matching is performed using the historical object matching points corresponding to each historical object image and the current object matching points corresponding to each current object image.

[0057] The same object point refers to the historical object matching points and the current object matching points being the same object point on the target object.

[0058] Specifically, the terminal randomly determines a target historical object matching point among the historical object matching points corresponding to the historical object images, respectively calculates similarities between the historical object matching points and current object matching points corresponding to the current object images to obtain similarity results corresponding to the current object matching points, determines a target current object matching point corresponding to the target historical object matching point among the current object matching points according to the similarity results, and the target historical object matching point and the target current object matching point are the same object point. The terminal traverses the historical object matching points and the current object matching points to obtain the historical object matching points and the current object matching points corresponding to each same object point.

[0059] In step 208, when the historical object matching point and the current feature matching point are the same object point, spatial pose transformation parameter calculation is performed using the historical object matching points corresponding to the historical object images to obtain historical spatial pose transformation parameters, and three-dimensional spatial matching point calculation is performed using the historical pose transformation parameters and the historical object matching points corresponding to the historical object images to obtain historical feature point three-dimensional coordinates.

[0060] In step 208, when the historical object matching point and the current feature matching point are the same object point, spatial pose transformation parameter calculation is performed using the historical object matching points corresponding to the historical object images to obtain historical spatial pose transformation parameters, and three-dimensional spatial matching point calculation is performed using the historical pose transformation parameters and the historical object matching points corresponding to the historical object images to obtain historical feature point three-dimensional coordinates.

[0061] Specifically, when the terminal detects the historical object matching points and the current object matching points corresponding to the same object point, the terminal obtains image coordinates of the historical object matching points corresponding to each same object point, and the historical object matching points can be a pair of matching points. The terminal performs spatial pose transformation parameter calculation according to the image coordinates of the historical object matching points corresponding to each same object point to obtain historical spatial pose transformation parameters. Then, the terminal performs three-dimensional spatial matching point calculation using the historical pose transformation parameters and the historical object matching points corresponding to each same object point to obtain historical feature point three-dimensional coordinates corresponding to each same object point.

[0062] In step 210, spatial pose transformation parameter calculation is performed using the current feature matching points corresponding to the current object images to obtain current spatial pose transformation parameters, and three-dimensional spatial matching point calculation is performed using the current pose transformation parameters and the current feature matching points corresponding to the current object images to obtain current feature point three-dimensional coordinates.

[0063] In step 210, spatial pose transformation parameter calculation is performed using the current feature matching points corresponding to the current object images to obtain current spatial pose transformation parameters, and three-dimensional spatial matching point calculation is performed using the current pose transformation parameters and the current feature matching points corresponding to the current object images to obtain current feature point three-dimensional coordinates.

[0064] Specifically, the terminal calculates the spatial pose transformation parameter according to the image coordinates of the current object matching points corresponding to each same object point, to obtain the current spatial pose transformation parameter. Then the terminal calculates the three-dimensional space matching points using the current pose transformation parameter and the current object matching points corresponding to each same object point, to obtain the current feature point three-dimensional coordinates corresponding to each same object point.

[0065] In step 212, the difference value corresponding to the historical feature point three-dimensional coordinates and the current feature point three-dimensional coordinates is calculated, to obtain the three-dimensional offset information corresponding to the target object.

[0066] The three-dimensional offset information refers to the information of the physical deformation of the target object.

[0067] Specifically, the terminal obtains the historical feature point three-dimensional coordinates and the current feature point three-dimensional coordinates corresponding to each same object point, respectively calculates the difference value corresponding to each same object point, calculates the average value according to the difference value corresponding to each same object point, and obtains the three-dimensional offset information corresponding to the target object.

[0068] In the above image processing method, the object point matching is respectively performed on the historical object image set and the current object image set, to obtain the historical object matching points corresponding to each historical object image in the historical object image set and the current object matching points corresponding to each current object image in the current object image set. When it is detected that the historical object matching point and the current feature matching point are the same object point, the corresponding spatial pose transformation parameter is calculated according to the historical object matching point and the current feature matching point, respectively. Then the historical feature point three-dimensional coordinates corresponding to the historical object matching point and the current feature point three-dimensional coordinates corresponding to the current feature matching point are calculated through the spatial pose transformation parameter, which realizes the conversion of the historical object matching point and the current feature matching point corresponding to the same object point from the image coordinates to the three-dimensional coordinates. The accuracy of the historical feature point three-dimensional coordinates and the current feature point three-dimensional coordinates is improved, so that the three-dimensional offset information corresponding to the target object calculated through the historical feature point three-dimensional coordinates and the current feature point three-dimensional coordinates is more accurate, and the accuracy of the deformation monitoring of the target object is improved.

[0069] In one embodiment, in step 204, the object point matching is respectively performed on the historical object image set and the current object image set, to obtain the historical object matching points corresponding to each historical object image in the historical object image set and the current object matching points corresponding to each current object image in the current object image set, including:

[0070] The blur processing is performed on each historical object image, to obtain the blur image corresponding to each historical object image, and the extreme value point is determined based on the pixel value in the blur image.

[0071] The object feature point corresponding to each historical object image is obtained based on the extreme value point.

[0072] extracting feature vectors corresponding to the object feature points, and calculating vector distances between the object feature points corresponding to each of the historical object images using the feature vectors;

[0073] performing feature matching on the object feature points corresponding to each of the historical object images based on the vector distances, to obtain historical object matching points corresponding to each of the historical object images.

[0074] The blurred image refers to a historical object image that has been subjected to blurring processing. The object feature point refers to a feature point on a target object in a historical object image.

[0075] Specifically, the terminal blurs each of the historical object images to obtain a blurred image corresponding to each of the historical object images. The terminal detects a change in pixel value in the blurred image, determines an extreme value point based on a point with the largest change in pixel value, and the extreme value point can be an edge pixel value of the target object. The terminal takes the extreme value point as an object feature point to obtain object feature points corresponding to each of the historical object images, and the object feature points can be at least three. Each of the historical object images can be a historical object image collected from the left side of the target object and a historical object image collected from the right side of the target object.

[0076] The terminal then extracts feature vectors of the object feature points, and calculates vector distances between the object feature points corresponding to each of the historical object images using the feature vectors. The terminal takes object feature points with the smallest vector distance in each of the historical object images as historical object matching points to obtain historical object matching points corresponding to each of the historical object images.

[0077] In one specific embodiment, the server can use a SIFT algorithm (Scale Invariant Feature Transform) to extract historical object matching points corresponding to each of the historical object images and current object matching points corresponding to each of the current object images. Each of the historical object images is specifically two historical object images collected from the left and right sides of the target object. The server takes the historical object image collected from the left side as a reference image and the historical object image collected from the right side as a target image.

[0078] The specific implementation steps of the SIFT algorithm include:

[0079] (1) Feature point detection. A Gaussian pyramid is constructed by using a Gaussian function to perform convolution operation on an image, and a Gaussian difference (DOG) pyramid is obtained by performing difference operation. In addition, precise positioning of key points is achieved by fitting a three-dimensional quadratic function to accurately determine the positions of the key points to achieve sub-pixel level accuracy.

[0080] (2) Determination of direction angle. After determining the feature points in each image in the previous step, it is necessary to calculate a direction for each feature point and then perform further calculations based on this direction. The principle is to use the gradient direction distribution characteristics of the local neighborhood pixels of the feature point to assign one or more direction angles to each feature point. All subsequent operations are based on the position, scale, and angle of the feature point. The direction angle of the feature point is constructed by using the statistical histogram of the gradient direction of each pixel in the neighborhood window of the feature point to construct the feature description vector. The feature description vector is shown in formula (1):

[0081]

[0082] Where L represents the pixel grayscale value at the corresponding point, and m(x,y) and θ(x,y) are the gradient magnitude and direction of the pixel (x,y), respectively. After performing the above calculations on 64 points, a histogram is used for statistical analysis. The horizontal axis of the histogram represents the gradient direction angle (0 to 360 degrees, with one bar for every 10 degrees), and the vertical axis represents the Gaussian weighted sum of the corresponding gradient values.

[0083] (3) Generating Feature Point Descriptors. After obtaining the principal direction and magnitude of the feature points, it is necessary to describe the feature points to prepare for matching between them. First, rotate the coordinate axes to the principal direction of the feature points. Only by describing the feature points with the principal direction as the zero point can they have rotation invariance. To enhance the robustness of matching, each feature point can be described using 4×4 seed points, a total of 16. Since each seed point has 8 direction vectors, each feature point can generate a 128-dimensional vector. These 128 values ​​are the SIFT feature point descriptors. At this point, the feature point description vectors are no longer affected by geometric factors such as scale changes and rotation. Finally, the length of the SIFT feature vectors is normalized to remove the influence of illumination changes.

[0084] (4) Feature point matching. After obtaining the feature point descriptors of the two images, the Euclidean distance between the feature vectors of the two feature points is used as the similarity criterion for the feature points in the two images. The Euclidean distance is calculated as shown in formula (2):

[0085]

[0086] Where xi and yi are the feature vector components of the feature points to be matched in the two images, respectively.

[0087] The server selects a feature point in the reference image and iterates through the target image to find the two feature points with the shortest Euclidean distance. If the ratio of the nearest distance to the second nearest distance is less than a preset distance threshold, the point is considered a matching point, and the matching points in the reference image and the target image are considered matching point pairs. The server iterates through the feature points in the reference and target images to determine each matching point pair, sorts them by Euclidean distance, and removes matching point pairs with larger distances to obtain the historical object matching points corresponding to each historical object image.

[0088] Repeat the above execution logic to obtain the current object matching point corresponding to each current object image.

[0089] In this embodiment, feature matching is performed on the object feature points corresponding to each historical object image by vector distance to obtain the historical object matching points corresponding to each historical object image in the historical object image set. This improves the accuracy of the historical object matching points, and thus the three-dimensional coordinates of the historical feature points calculated by the historical object matching points are more accurate, thereby improving the accuracy of the three-dimensional offset information of the target object.

[0090] In one embodiment, step 206, matching the same object point using the historical object matching points corresponding to each historical object image and the current object matching points corresponding to each current object image, includes:

[0091] Extract the feature vectors corresponding to historical object matching points and the feature vectors corresponding to the current object matching point;

[0092] Calculate the vector distance between the feature vector corresponding to the historical object matching point and the feature vector corresponding to the current object matching point;

[0093] When the vector distance is detected to be less than the preset distance threshold, the historical object matching point and the current object matching point are determined to be the same object matching point.

[0094] Specifically, the server extracts the feature vectors corresponding to each historical object matching point and the feature vectors corresponding to each current object matching point. The vector distance between the feature vectors corresponding to each historical object matching point and the feature vectors corresponding to each current object matching point is calculated. Historical object matching points and current object matching points whose vector distance is less than a preset distance threshold are identified as the same object matching point, thus obtaining the same object matching points between the historical object image and the current object image.

[0095] In this embodiment, by determining whether the historical object matching point and the current object matching point are the same object matching point based on the vector distance, the mismatch between the historical object matching point and the current object matching point is reduced, thereby improving the accuracy of the coordinate difference corresponding to each point of the same object.

[0096] In one embodiment, step 208 involves calculating spatial pose transformation parameters using the historical object matching points corresponding to each historical object image, to obtain the historical spatial pose transformation parameters, including:

[0097] Obtain the image coordinates of the historical object matching points corresponding to each historical object image, and calculate the coordinate mapping relationship parameters between each historical object image based on the image coordinates.

[0098] Based on the coordinate mapping relationship parameters, parameter decomposition is performed to obtain the historical spatial pose transformation parameters.

[0099] Among them, the coordinate mapping relationship parameter refers to the parameter that characterizes the coordinate mapping relationship between various historical object images.

[0100] Specifically, the server acquires the image coordinates of the historical object matching points corresponding to each historical object image, and calculates the coordinate mapping relationship parameters between the historical object images based on the image coordinates. These coordinate mapping relationship parameters can be the basic matrices corresponding to each historical object image. The server performs parametric decomposition on the basic matrices to obtain historical spatial pose transformation parameters. These historical spatial pose transformation parameters can be the pose transformation parameters of the image acquisition device when acquiring images from the left and right sides of the target object, and are used to calculate the three-dimensional coordinates of the historical feature points corresponding to the historical object matching points.

[0101] By repeating the above execution logic, the server obtains the current spatial pose transformation parameters corresponding to each historical object image.

[0102] In one specific embodiment, the fundamental matrix corresponding to each historical object image is calculated using a fundamental matrix estimation method. The fundamental matrix is ​​defined by equation (3):

[0103] X'FX = 0 Formula (3)

[0104] Among them, It is any pair of matching points between two images.

[0105] Since the matching of each set of points provides a linear equation for calculating the F coefficients, given at least 7 points (a 3×3 homogeneous matrix minus a scale, and a rank-2 constraint), the equation can calculate the unknown F, i.e., the coordinate mapping parameters. We denote the coordinates of the points as X = (x, y, 1). T X' = ​​(x', y', 1) T The corresponding equation is shown in formula (4):

[0106]

[0107] The expanded form is shown in formula (5):

[0108] x'xf11 +x'yf 12 +x'f 13 +y'xf 21 +y'yf 22 +y'f 23 +xf 11 +yf 22 +f 33 =0

[0109] Formula (5)

[0110] If matrix F is written in column vector form, it is as shown in formula (6):

[0111] [x'x x'y x' y'x y'y y' xy 1]f=0 Formula (6)

[0112] Given a set of n points, the result is as shown in formula (7):

[0113]

[0114] The algorithm for determining the fundamental matrix includes:

[0115] (1) Normalization: According to Transform the image coordinates, where T and T' are normalized transformations achieved by translation and scaling.

[0116] (2) Solve for the basic matrix of corresponding matching include:

[0117] Find the linear solution: using the corresponding point set Determined coefficient matrix Determination of the singular vector of the minimum singular value

[0118] Singularity constraints: Use SVD (Singular Value Decomposition) to... Decompose it and set its minimum singular value to 0 to obtain Make

[0119] (3) Remove normalization: Let Matrix F is a point The corresponding basic matrix.

[0120] The server calculates the basic matrix corresponding to each historical object image as follows: server Singular value decomposition yields the historical spatial pose transformation parameters as follows: Where R represents the rotation matrix and t represents the translation matrix.

[0121] In this embodiment, the coordinate mapping relationship parameters between various historical object images are calculated based on the image coordinates of the historical object matching points, and the coordinate mapping relationship parameters are decomposed to obtain the historical spatial pose transformation parameters. This improves the accuracy of the historical feature point 3D coordinates by using the historical spatial pose transformation parameters to calculate the historical feature point 3D coordinates.

[0122] In one embodiment, step 208 involves using a camera device to calculate three-dimensional spatial matching points using historical pose transformation parameters and historical object matching points corresponding to each historical object image, thereby obtaining the three-dimensional coordinates of historical feature points, including:

[0123] Obtain the calibration parameters corresponding to the camera device, and calculate the three-dimensional spatial matching points based on the calibration parameters, historical pose transformation parameters and matching points of each historical object to obtain the three-dimensional coordinates of the historical feature points.

[0124] Among them, calibration parameters refer to the inherent attribute parameters of the camera device, which are used to calculate matching points in three-dimensional space.

[0125] Specifically, the server obtains the calibration parameters corresponding to the camera device, which can be the intrinsic parameters of the camera device. Based on the calibration parameters, historical pose transformation parameters, and matching points of each historical object, the server calculates the 3D spatial matching points to obtain the 3D coordinates of the historical feature points. The server can use triangulation methods to calculate the 3D coordinates of the historical feature points.

[0126] The triangulation calculation is shown in formula (8), which represents the projection of a point in three-dimensional space onto the image.

[0127]

[0128] Where λ represents the camera depth; K represents the calibration parameter, i.e., the intrinsic parameter; a = [uv 1] T This represents the normalized planar coordinates, obtained by normalizing pixel coordinates; P represents the pose transformation parameters of the camera device; X represents the homogeneous coordinates of a point in 3D space [xyz 1]. T P1, P2, and P3 represent the pose transformation parameters from the left camera to the right camera, the pose transformation parameters from the right camera to the left camera, and the pose transformation parameters between the left and right cameras, respectively.

[0129] The cross product of both sides of formula (8) is PX, as shown in formula (9):

[0130]

[0131]

[0132] in, Let X represent the antisymmetric matrix corresponding to vector a. The server uses the SVD algorithm to calculate the singular vector of the minimum singular value of the homography matrix corresponding to formula (9), and obtains the homogeneous coordinates X, that is, the three-dimensional coordinates of the historical feature points.

[0133] Repeat the above execution logic to obtain the 3D coordinates of the current feature point corresponding to the current object matching point in each current object image.

[0134] In this embodiment, by using calibration parameters, historical pose transformation parameters, and matching points of each historical object to calculate the three-dimensional spatial matching points, the accuracy of the three-dimensional coordinates of historical feature points is improved. This allows for the subsequent use of the three-dimensional coordinates of historical feature points to calculate the three-dimensional offset information corresponding to the target object, thereby improving the accuracy of the three-dimensional offset information.

[0135] In one embodiment, step 212, calculating the difference between the historical 3D coordinates of the feature point and the current 3D coordinates of the feature point to obtain the 3D offset information corresponding to the target object, includes:

[0136] Obtain the 3D coordinates of historical feature points and the current feature point for each point of the same object. Calculate the difference between the 3D coordinates of historical feature points and the current feature point for each point of the same object to obtain the coordinate difference for each point of the same object.

[0137] The three-dimensional offset information of the target object is obtained by averaging the coordinate differences of each point on the same object.

[0138] Specifically, the server obtains the 3D coordinates of historical feature points and the current 3D coordinates of the same object point, and calculates the difference between the 3D coordinates of historical feature points and the current 3D coordinates of the same object point to obtain the coordinate difference of the same object point. The difference calculation is shown in formula (10):

[0139]

[0140] Where Δx, Δy, and Δz represent the three-dimensional differences between the historical 3D coordinates of a feature point and the current 3D coordinates of a feature point. d represents the coordinate difference between points on the same object.

[0141] The server filters the coordinate differences corresponding to various points of the same object, eliminating outliers to obtain the coordinate differences of each target. The server then calculates the average of these coordinate differences to obtain the 3D offset information of the target object.

[0142] In one specific embodiment, such as Figure 3The diagram illustrates a process for acquiring 3D offset information. First, the camera is calibrated to obtain calibration parameters. Then, a set of historical object images and a set of current object images are acquired. Feature extraction and feature matching are performed on both sets to obtain historical object matching points and current object matching points. The server uses the historical object matching points to estimate the fundamental matrix between historical object images, calculates the historical spatial pose transformation parameters corresponding to the historical object matching points based on this matrix, and uses the current object matching points to estimate the fundamental matrix between current object images, calculates the current spatial pose transformation parameters corresponding to the current object matching points based on this matrix.

[0143] The server uses the pixel coordinates, calibration parameters, and historical spatial pose transformation parameters of the historical object matching points to calculate the 3D coordinates of the historical feature points corresponding to the historical object matching points using a triangulation method. Similarly, it uses the pixel coordinates, calibration parameters, and current spatial pose transformation parameters of the current object matching point to calculate the 3D coordinates of the current feature points corresponding to the current object matching points using a triangulation method.

[0144] The server calculates the 3D offset information of the target object based on the historical 3D coordinates of the feature points and the current 3D coordinates of the feature points.

[0145] In one specific embodiment, such as Figure 4 The diagram illustrates a specific process for obtaining 3D offset information. The server acquires a set of historical object images and a set of current object images. It then performs Gaussian blurring, constructs a Gaussian pyramid, constructs a Gaussian subdivision pyramid, searches for extreme points, precisely locates key points, determines the main direction of key points, and generates feature point descriptors—that is, the object feature points corresponding to each historical object image in the historical image set and the object feature points corresponding to each current object image in the current object image set.

[0146] The server iterates through the Euclidean distances of image feature points corresponding to each historical object image and the Euclidean distances of image feature points corresponding to each current object image. The server sorts the matching points based on their Euclidean distances, obtaining the historical and current object matching points with the shorter Euclidean distances. The server then extracts the pixel coordinates of the historical and current object matching points and calculates the 3D offset information of the target object based on these coordinates.

[0147] In one specific embodiment, such as Figure 5 As shown, a triangulation diagram is provided. In the diagram, p... j P represents the feature points on the object image acquired by the image acquisition device.j This represents the 3D spatial matching point corresponding to the feature point. K represents the calibration parameter corresponding to the image acquisition device.

[0148] It should be understood that although the steps in the flowcharts of the embodiments described above are shown sequentially according to the arrows, these steps are not necessarily executed in the order indicated by the arrows. Unless explicitly stated herein, there is no strict order restriction on the execution of these steps, and they can be executed in other orders. Moreover, at least some steps in the flowcharts of the embodiments described above may include multiple steps or multiple stages. These steps or stages are not necessarily completed at the same time, but can be executed at different times. The execution order of these steps or stages is not necessarily sequential, but can be performed alternately or in turn with other steps or at least some of the steps or stages of other steps.

[0149] Based on the same inventive concept, this application also provides an image processing apparatus for implementing the image processing method described above. The solution provided by this apparatus is similar to the implementation scheme described in the above method; therefore, the specific limitations in one or more image processing apparatus embodiments provided below can be found in the limitations of the image processing method described above, and will not be repeated here.

[0150] In one embodiment, such as Figure 6 As shown, an image processing apparatus 600 is provided, including: an acquisition module 602, a matching module 604, a matching point recognition module 606, a historical coordinate transformation module 608, a current coordinate transformation module 610, and an offset calculation module 612, wherein:

[0151] The acquisition module 602 is used to acquire the historical object image set and the current object image set corresponding to the target object. The historical object image set is obtained by the image acquisition device acquiring images of the target object from different acquisition positions at historical times. The current object image set is obtained by the image acquisition device acquiring images of the target object from different acquisition positions at the current time.

[0152] The matching module 604 is used to perform object point matching based on the historical object image set and the current object image set respectively, to obtain the historical object matching points corresponding to each historical object image in the historical object image set and the current object matching points corresponding to each current object image in the current object image set.

[0153] The matching point recognition module 606 is used to perform the same object point matching using the historical object matching points corresponding to each historical object image and the current object matching points corresponding to each current object image.

[0154] The historical coordinate transformation module 608 is used to calculate the spatial pose transformation parameters using the historical object matching points corresponding to each historical object image when the historical object matching point and the current feature matching point are the same object point, and to calculate the three-dimensional spatial matching point using the historical pose transformation parameters and the historical object matching points corresponding to each historical object image, so as to obtain the three-dimensional coordinates of the historical feature point.

[0155] The current coordinate transformation module 610 is used to calculate the spatial pose transformation parameters using the current feature matching points corresponding to each current object image, to obtain the current spatial pose transformation parameters, and to calculate the three-dimensional spatial matching points using the current pose transformation parameters and the current feature matching points corresponding to each current object image, to obtain the three-dimensional coordinates of the current feature points.

[0156] The offset calculation module 612 is used to calculate the difference between the three-dimensional coordinates of historical feature points and the three-dimensional coordinates of current feature points to obtain the three-dimensional offset information of the target object.

[0157] In one embodiment, the matching module 604 includes:

[0158] The feature matching unit is used to blur each historical object image to obtain a blurred image corresponding to each historical object image, and to determine the extreme points based on the pixel values ​​in the blurred image; to obtain the object feature points corresponding to each historical object image based on the extreme points; to extract the feature vectors corresponding to the object feature points, and to use the feature vectors to calculate the vector distance between the object feature points corresponding to each historical object image; and to perform feature matching on the object feature points corresponding to each historical object image based on the vector distance to obtain the historical object matching points corresponding to each historical object image in the historical object image set.

[0159] In one embodiment, the matching point recognition module 606 includes:

[0160] The matching point detection unit is used to extract the feature vectors corresponding to historical object matching points and the feature vectors corresponding to current object matching points; calculate the vector distance between the feature vectors corresponding to historical object matching points and the feature vectors corresponding to current object matching points; and determine that the historical object matching point and the current object matching point are the same object matching point when the vector distance is less than a preset distance threshold.

[0161] In one embodiment, the historical coordinate transformation module 608 includes:

[0162] The parameter decomposition unit is used to obtain the image coordinates of the historical object matching points corresponding to each historical object image, calculate the coordinate mapping relationship parameters between each historical object image based on the image coordinates, and perform parameter decomposition based on the coordinate mapping relationship parameters to obtain the historical spatial pose transformation parameters.

[0163] In one embodiment, the historical coordinate transformation module 608 includes:

[0164] The 3D point calculation unit is used to obtain the calibration parameters corresponding to the camera device, and to calculate the 3D spatial matching points based on the calibration parameters, historical pose transformation parameters and matching points of each historical object, so as to obtain the 3D coordinates of the historical feature points.

[0165] In one embodiment, the offset calculation module 612 includes:

[0166] The mean calculation unit is used to obtain the three-dimensional coordinates of historical feature points and the current three-dimensional coordinates of feature points corresponding to each point of the same object, calculate the difference between the three-dimensional coordinates of historical feature points and the current three-dimensional coordinates of feature points corresponding to each point of the same object, and obtain the coordinate difference corresponding to each point of the same object; based on the coordinate difference corresponding to each point of the same object, the mean is calculated to obtain the three-dimensional offset information corresponding to the target object.

[0167] Each module in the aforementioned image processing device can be implemented entirely or partially through software, hardware, or a combination thereof. These modules can be embedded in or independent of the processor in a computer device, or stored in the memory of a computer device as software, so that the processor can call and execute the operations corresponding to each module.

[0168] In one embodiment, a computer device is provided, which may be a server, and its internal structure diagram may be as follows: Figure 7 As shown, this computer device includes a processor, memory, input / output (I / O) interfaces, and a communication interface. The processor, memory, and I / O interfaces are connected via a system bus, and the communication interface is also connected to the system bus via the I / O interfaces. The processor provides computational and control capabilities. The memory includes non-volatile storage media and internal memory. The non-volatile storage media stores the operating system, computer programs, and a database. The internal memory provides the environment for the operating system and computer programs stored in the non-volatile storage media. The database stores historical and current sets of object images. The I / O interfaces allow the processor to exchange information with external devices. The communication interface allows communication with external terminals via a network connection. When executed by the processor, the computer program implements an image processing method.

[0169] In one embodiment, a computer device is provided, which may be a terminal, and its internal structure diagram may be as follows: Figure 8As shown, the computer device includes a processor, memory, input / output interfaces, a communication interface, a display unit, and an input device. The processor, memory, and input / output interfaces are connected via a system bus, and the communication interface, display unit, and input device are also connected to the system bus via the input / output interfaces. The processor provides computational and control capabilities. The memory includes a non-volatile storage medium and internal memory. The non-volatile storage medium stores the operating system and computer programs. The internal memory provides an environment for the operation of the operating system and computer programs stored in the non-volatile storage medium. The input / output interfaces are used for the processor to exchange information with external devices. The communication interface is used for wired or wireless communication with external terminals; wireless communication can be achieved through Wi-Fi, mobile cellular networks, NFC (Near Field Communication), or other technologies. When the computer program is executed by the processor, it implements an image processing method. The display unit of the computer device is used to form a visually visible image. It can be a display screen, a projection device, or a virtual reality imaging device. The display screen can be an LCD screen or an e-ink screen. The input device of the computer device can be a touch layer covering the display screen, or buttons, trackballs, or touchpads set on the casing of the computer device, or external keyboards, touchpads, or mice, etc.

[0170] Those skilled in the art will understand that Figures 7-8 The structure shown is merely a block diagram of a portion of the structure related to the present application and does not constitute a limitation on the computer device to which the present application is applied. Specific computer devices may include more or fewer components than those shown in the figure, or combine certain components, or have different component arrangements.

[0171] In one embodiment, a computer device is also provided, including a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to implement the steps in the above method embodiments.

[0172] In one embodiment, a computer-readable storage medium is provided having a computer program stored thereon that, when executed by a processor, implements the steps in the above method embodiments.

[0173] In one embodiment, a computer program product is provided, including a computer program that, when executed by a processor, implements the steps in the above method embodiments.

[0174] It should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for analysis, data stored, data displayed, etc.) involved in this application are all information and data authorized by the user or fully authorized by all parties, and the collection, use and processing of the relevant data shall comply with the relevant laws, regulations and standards of the relevant countries and regions.

[0175] Those skilled in the art will understand that all or part of the processes in the methods of the above embodiments can be implemented by a computer program instructing related hardware. The computer program can be stored in a non-volatile computer-readable storage medium, and when executed, it can include the processes of the embodiments of the above methods. Any references to memory, databases, or other media used in the embodiments provided in this application can include at least one of non-volatile and volatile memory. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory, optical memory, high-density embedded non-volatile memory, resistive random access memory (ReRAM), magnetic random access memory (MRAM), ferroelectric random access memory (FRAM), phase change memory (PCM), graphene memory, etc. Volatile memory can include random access memory (RAM) or external cache memory, etc. By way of illustration and not limitation, RAM can take many forms, such as Static Random Access Memory (SRAM) or Dynamic Random Access Memory (DRAM). The databases involved in the embodiments provided in this application may include at least one type of relational database and non-relational database. Non-relational databases may include, but are not limited to, blockchain-based distributed databases. The processors involved in the embodiments provided in this application may be general-purpose processors, central processing units, graphics processing units, digital signal processors, programmable logic devices, quantum computing-based data processing logic devices, etc., and are not limited to these.

[0176] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0177] The embodiments described above are merely illustrative of several implementation methods of this application, and while the descriptions are specific and detailed, they should not be construed as limiting the scope of this patent application. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of this application, and these all fall within the protection scope of this application. Therefore, the protection scope of this application should be determined by the appended claims.

Claims

1. An image processing method, characterized in that, The method includes: The system acquires a historical image set and a current image set corresponding to the target object. The historical image set is obtained by the image acquisition device capturing images of the target object from different acquisition positions at historical times. The current image set is obtained by the image acquisition device capturing images of the target object from different acquisition positions at the current time. Based on the historical object image set and the current object image set, object point matching is performed to obtain the historical object matching point corresponding to each historical object image in the historical object image set and the current object matching point corresponding to each current object image in the current object image set. The same object point is matched using the historical object matching points corresponding to each historical object image and the current object matching points corresponding to each current object image. When the historical object matching point and the current object matching point are the same object point, the spatial pose transformation parameters are calculated using the historical object matching points corresponding to each historical object image to obtain the historical spatial pose transformation parameters. Then, the three-dimensional spatial matching point is calculated using the historical spatial pose transformation parameters and the historical object matching points corresponding to each historical object image to obtain the three-dimensional coordinates of the historical feature point. Spatial pose transformation parameters are calculated using the current feature matching points corresponding to each current object image to obtain the current spatial pose transformation parameters. Then, three-dimensional spatial matching point calculation is performed using the current spatial pose transformation parameters and the current feature matching points corresponding to each current object image to obtain the three-dimensional coordinates of the current feature points. The difference between the three-dimensional coordinates of the historical feature point and the three-dimensional coordinates of the current feature point is calculated to obtain the three-dimensional offset information of the target object.

2. The method according to claim 1, characterized in that, The step of performing object point matching based on the historical object image set and the current object image set to obtain historical object matching points corresponding to each historical object image in the historical object image set and current object matching points corresponding to each current object image in the current object image set includes: The images of each historical object are blurred to obtain blurred images corresponding to each historical object, and extreme points are determined based on the pixel values ​​in the blurred images. Based on the extreme points, the object feature points corresponding to each historical object image are obtained respectively; Extract the feature vectors corresponding to the feature points of the object, and use the feature vectors to calculate the vector distance between the feature points of the object corresponding to each historical object image; Based on the vector distance, feature matching is performed on the object feature points corresponding to each historical object image to obtain the historical object matching points corresponding to each historical object image in the historical object image set.

3. The method according to claim 1, characterized in that, The step of matching the same object point using the historical object matching points corresponding to each historical object image and the current object matching points corresponding to each current object image includes: Extract the feature vectors corresponding to the historical object matching points and the feature vectors corresponding to the current object matching points; Calculate the vector distance between the feature vector corresponding to the historical object matching point and the feature vector corresponding to the current object matching point; When the vector distance is detected to be less than a preset distance threshold, the historical object matching point and the current object matching point are determined to be the same object matching point.

4. The method according to claim 1, characterized in that, The step of using the historical object matching points corresponding to each historical object image to calculate the spatial pose transformation parameters, and obtaining the historical spatial pose transformation parameters, includes: Obtain the image coordinates of the historical object matching points corresponding to each historical object image, and calculate the coordinate mapping relationship parameters between each historical object image based on the image coordinates; Based on the coordinate mapping relationship parameters, parameter decomposition is performed to obtain the historical spatial pose transformation parameters.

5. The method according to claim 1, characterized in that, The image acquisition device is a camera device; the step of using the historical spatial pose transformation parameters and the historical object matching points corresponding to each historical object image to calculate the three-dimensional spatial matching points and obtain the three-dimensional coordinates of the historical feature points includes: Obtain the calibration parameters corresponding to the camera device, and calculate the three-dimensional spatial matching points based on the calibration parameters, the historical pose transformation parameters, and the matching points of each historical object to obtain the three-dimensional coordinates of the historical feature points.

6. The method according to claim 1, characterized in that, The step of calculating the difference between the three-dimensional coordinates of the historical feature point and the three-dimensional coordinates of the current feature point to obtain the three-dimensional offset information corresponding to the target object includes: Obtain the 3D coordinates of historical feature points and the 3D coordinates of current feature points corresponding to each point of the same object; calculate the difference between the 3D coordinates of historical feature points and the 3D coordinates of current feature points corresponding to each point of the same object to obtain the coordinate difference between each point of the same object. The average value is calculated based on the coordinate differences of the corresponding points of the same object to obtain the three-dimensional offset information of the target object.

7. An image processing apparatus, characterized in that, The device includes: The acquisition module is used to acquire a historical object image set and a current object image set corresponding to the target object. The historical object image set is obtained by the image acquisition device acquiring images of the target object from different acquisition positions at historical times. The current object image set is obtained by the image acquisition device acquiring images of the target object from different acquisition positions at the current time. The matching module is used to perform object point matching based on the historical object image set and the current object image set respectively, to obtain the historical object matching points corresponding to each historical object image in the historical object image set and the current object matching points corresponding to each current object image in the current object image set. The matching point recognition module is used to perform the same object point matching using the historical object matching points corresponding to each historical object image and the current object matching points corresponding to each current object image. The historical coordinate transformation module is used to calculate the spatial pose transformation parameters using the historical object matching points corresponding to each historical object image when the historical object matching point and the current object matching point are the same object point, and to calculate the three-dimensional spatial matching point using the historical spatial pose transformation parameters and the historical object matching points corresponding to each historical object image, so as to obtain the three-dimensional coordinates of the historical feature point. The current coordinate transformation module is used to calculate the spatial pose transformation parameters using the current feature matching points corresponding to each current object image, to obtain the current spatial pose transformation parameters, and to calculate the three-dimensional spatial matching points using the current spatial pose transformation parameters and the current feature matching points corresponding to each current object image, to obtain the three-dimensional coordinates of the current feature points. The offset calculation module is used to calculate the difference between the three-dimensional coordinates of the historical feature point and the three-dimensional coordinates of the current feature point to obtain the three-dimensional offset information of the target object.

8. A computer device comprising a memory and a processor, wherein the memory stores a computer program, characterized in that, When the processor executes the computer program, it implements the steps of the method according to any one of claims 1 to 6.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by a processor, it implements the steps of the method according to any one of claims 1 to 6.

10. A computer program product, comprising a computer program, characterized in that, When the computer program is executed by a processor, it implements the steps of the method according to any one of claims 1 to 6.

Images