CN112749709A - Image processing method and device, electronic equipment and storage medium - Google Patents
Image processing method and device, electronic equipment and storage medium Download PDFInfo
- Publication number
- CN112749709A CN112749709A CN201911045867.8A CN201911045867A CN112749709A CN 112749709 A CN112749709 A CN 112749709A CN 201911045867 A CN201911045867 A CN 201911045867A CN 112749709 A CN112749709 A CN 112749709A
- Authority
- CN
- China
- Prior art keywords
- group
- image
- pixel point
- processed
- feature map
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Withdrawn
Links
- 238000003860 storage Methods 0.000 title claims abstract description 24
- 238000003672 processing method Methods 0.000 title claims abstract description 14
- 230000009466 transformation Effects 0.000 claims abstract description 101
- 238000000034 method Methods 0.000 claims abstract description 45
- 239000013598 vector Substances 0.000 claims abstract description 24
- 230000001131 transforming effect Effects 0.000 claims abstract description 9
- 238000012545 processing Methods 0.000 claims description 77
- 238000000605 extraction Methods 0.000 claims description 37
- 238000011176 pooling Methods 0.000 claims description 21
- 238000004590 computer program Methods 0.000 claims description 15
- 238000013528 artificial neural network Methods 0.000 claims description 12
- 238000004364 calculation method Methods 0.000 abstract description 4
- 238000010586 diagram Methods 0.000 description 23
- 230000008859 change Effects 0.000 description 22
- 238000013527 convolutional neural network Methods 0.000 description 12
- 238000004891 communication Methods 0.000 description 10
- 238000005516 engineering process Methods 0.000 description 9
- 230000006870 function Effects 0.000 description 9
- 239000011159 matrix material Substances 0.000 description 7
- 238000006243 chemical reaction Methods 0.000 description 6
- 230000000007 visual effect Effects 0.000 description 6
- 230000003287 optical effect Effects 0.000 description 5
- 230000008569 process Effects 0.000 description 5
- 230000005540 biological transmission Effects 0.000 description 4
- 230000005236 sound signal Effects 0.000 description 4
- 230000001133 acceleration Effects 0.000 description 2
- 239000000835 fiber Substances 0.000 description 2
- 230000003993 interaction Effects 0.000 description 2
- 230000001902 propagating effect Effects 0.000 description 2
- 238000005070 sampling Methods 0.000 description 2
- 230000003068 static effect Effects 0.000 description 2
- RYGMFSIKBFXOCR-UHFFFAOYSA-N Copper Chemical compound [Cu] RYGMFSIKBFXOCR-UHFFFAOYSA-N 0.000 description 1
- 230000009471 action Effects 0.000 description 1
- 238000003491 array Methods 0.000 description 1
- 230000001413 cellular effect Effects 0.000 description 1
- 229910052802 copper Inorganic materials 0.000 description 1
- 239000010949 copper Substances 0.000 description 1
- 238000005286 illumination Methods 0.000 description 1
- 238000003384 imaging method Methods 0.000 description 1
- 239000004973 liquid crystal related substance Substances 0.000 description 1
- 238000004519 manufacturing process Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 230000002093 peripheral effect Effects 0.000 description 1
- 238000013341 scale-up Methods 0.000 description 1
- 239000004065 semiconductor Substances 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/40—Extraction of image or video features
- G06V10/46—Descriptors for shape, contour or point-related descriptors, e.g. scale invariant feature transform [SIFT] or bags of words [BoW]; Salient regional features
- G06V10/462—Salient features, e.g. scale invariant feature transforms [SIFT]
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T3/00—Geometric image transformations in the plane of the image
- G06T3/40—Scaling of whole images or parts thereof, e.g. expanding or contracting
- G06T3/4023—Scaling of whole images or parts thereof, e.g. expanding or contracting based on decimating pixels or lines of pixels; based on inserting pixels or lines of pixels
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T3/00—Geometric image transformations in the plane of the image
- G06T3/60—Rotation of whole images or parts thereof
- G06T3/604—Rotation of whole images or parts thereof using coordinate rotation digital computer [CORDIC] devices
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Multimedia (AREA)
- Image Analysis (AREA)
Abstract
The present disclosure relates to an image processing method and apparatus, an electronic device, and a storage medium, the method including: transforming the image to be processed according to at least one group of image transformation parameters to obtain a plurality of transformed images to be processed; respectively extracting image features of the transformed images to be processed to obtain a first group feature map corresponding to each pixel point in the images to be processed, wherein the first group feature map corresponding to any pixel point is used for representing a feature vector of the pixel point corresponding to the transformed images to be processed; extracting group structure information of the first group feature map corresponding to any pixel point to obtain a second group feature map corresponding to the pixel point; and obtaining the image descriptor of the image to be processed according to the second group feature map corresponding to each pixel point. The embodiment of the disclosure can improve the accuracy of the image descriptor and the robustness of scale and rotation transformation, and improve the calculation efficiency of extracting the image descriptor.
Description
Technical Field
The present disclosure relates to the field of computer technologies, and in particular, to an image processing method and apparatus, an electronic device, and a storage medium.
Background
Constructing a feature descriptor of an image is a fundamental problem in computer vision-based tasks, such as: the method comprises the steps of establishing an image descriptor which has strong expression capability and can resist visual angle difference and illumination difference in tasks such as three-dimensional reconstruction, visual positioning, image splicing, image retrieval and the like.
The robustness, accuracy and extraction efficiency of the image descriptor, which is important data for describing the image features, have an important influence on the overall performance of a task to which the image descriptor is applied.
Disclosure of Invention
The present disclosure proposes an image processing technical solution.
According to an aspect of the present disclosure, there is provided an image processing method including:
transforming the image to be processed according to at least one group of image transformation parameters to obtain a plurality of transformed images to be processed;
respectively extracting image features of the transformed images to be processed to obtain a first group feature map corresponding to each pixel point in the images to be processed, wherein the first group feature map corresponding to any pixel point is used for representing a feature vector of the pixel point corresponding to the transformed images to be processed;
extracting group structure information of the first group feature map corresponding to any pixel point to obtain a second group feature map corresponding to the pixel point;
and obtaining the image descriptor of the image to be processed according to the second group feature map corresponding to each pixel point.
In one possible implementation, the method further includes:
determining unit rotation amount and unit zooming amount;
determining at least one rotation angle according to the unit rotation amount and at least one rotation coefficient, and determining at least one scaling scale according to the unit scaling amount and at least one scaling coefficient;
and combining at least one group of image transformation parameters by using the at least one rotation angle and the at least one scaling scale, wherein each group of image transformation parameters comprises a rotation angle and a scaling scale.
In this way, a uniform sampling of the image variation parameters can be achieved.
In a possible implementation manner, the transforming the image to be processed according to at least one group of image transformation parameters to obtain a plurality of transformed images to be processed includes:
and for each group of image change parameters, rotating the image to be processed according to the rotation angle contained in the image change parameters, and scaling the image to be processed according to the scaling scale contained in the image change parameters to obtain the transformed image to be processed.
Therefore, a plurality of converted images to be processed can be obtained, the image descriptors of the images to be processed can be inferred by utilizing the characteristic conversion of the images to be processed under a plurality of groups of image conversion parameters and utilizing larger information quantity, and the accuracy of the image descriptors and the robustness of scale and rotation conversion can be improved.
In a possible implementation manner, the performing image feature extraction on the plurality of transformed images to be processed respectively to obtain a first group feature map corresponding to each pixel point in the images to be processed includes:
respectively carrying out image feature extraction on the plurality of transformed images to be processed to obtain feature maps corresponding to the plurality of transformed images to be processed;
and aiming at any pixel point, obtaining a first group of feature maps of the pixel point according to the feature vectors of the pixel point in the feature maps corresponding to the transformed images to be processed.
Therefore, the feature transformation of the image to be processed under a plurality of groups of image transformation parameters is utilized, the first group feature map of each pixel is obtained through a plurality of transformed images to be processed, the image descriptor of the image to be processed is inferred by utilizing larger information quantity, and the accuracy of the image descriptor and the robustness of scale and rotation transformation can be improved.
In a possible implementation manner, the extracting group structure information of the first group feature map corresponding to any pixel point to obtain a second group feature map corresponding to the pixel point includes:
and performing group equal variation convolution processing on the first group feature graph corresponding to any pixel point to obtain a second group feature graph corresponding to the pixel point, wherein the second group feature graph comprises group structure information of the first group feature graph.
Therefore, by carrying out group convolution processing on the first group characteristic diagram, the learned image descriptor can resist geometric transformation caused by visual angle change, the robustness is improved, and the accuracy of the image descriptor can be improved.
In a possible implementation manner, the performing group invariant convolution processing on the first group feature map corresponding to any pixel point to obtain a second group feature map corresponding to the pixel point includes:
for each group of image transformation parameters, determining a neighborhood of the image transformation parameter from each group of image transformation parameters;
determining a plurality of convolution regions in the first group of feature maps according to the neighborhood corresponding to each image transformation parameter aiming at any pixel point;
and performing convolution operation on the plurality of convolution areas to obtain a second group feature map corresponding to the pixel point.
Therefore, the learned image descriptor can resist the geometric transformation caused by the change of the visual angle, the robustness is improved, and the accuracy of the image descriptor can be improved.
In a possible implementation manner, the obtaining an image descriptor of the image to be processed according to the second group feature map corresponding to each pixel point includes:
and performing pooling processing on the second group feature map to obtain an image descriptor of the image to be processed.
In a possible implementation manner, the extracting group structure information of the first group feature map corresponding to any pixel point to obtain a second group feature map corresponding to the pixel point includes:
performing twice group equal variation convolution processing on the first group feature graph corresponding to any pixel point to obtain a first sub-graph and a second sub-graph corresponding to the pixel point, wherein the first sub-graph corresponds to a result of the first group equal variation convolution processing, and the second sub-graph corresponds to a result of the second group equal variation convolution processing;
the obtaining the image descriptor of the image to be processed according to the second group feature map corresponding to each pixel point includes:
and carrying out bilinear interpolation pooling processing according to the first subgraph and the second subgraph to obtain an image descriptor of the image to be processed.
In this way, the second group feature map is subjected to pooling processing in a bilinear interpolation pooling mode, and the image descriptor with stronger expression capability can be obtained.
In one possible implementation, the method is implemented by a neural network.
According to an aspect of the present disclosure, there is provided an image processing apparatus including:
the transformation module is used for transforming the image to be processed according to at least one group of image transformation parameters to obtain a plurality of transformed images to be processed;
the first feature extraction module is used for respectively extracting image features of the plurality of transformed images to be processed to obtain a first group feature map corresponding to each pixel point in the images to be processed, wherein the first group feature map corresponding to any pixel point is used for representing a feature vector of the pixel point corresponding to the plurality of transformed images to be processed;
the second feature extraction module is used for extracting group structure information of the first group feature map corresponding to any pixel point to obtain a second group feature map corresponding to the pixel point;
and the processing module is used for obtaining the image descriptor of the image to be processed according to the second group characteristic graph corresponding to each pixel point.
In one possible implementation, the apparatus further includes:
the first determining module is used for determining unit rotation amount and unit zooming amount;
the second determining module is used for determining at least one rotation angle according to the unit rotation amount and at least one rotation coefficient, and determining at least one scaling scale according to the unit scaling amount and at least one scaling coefficient;
and the combination module is used for combining at least one group of image transformation parameters by using the at least one rotation angle and the at least one scaling scale, wherein each group of image transformation parameters comprises a rotation angle and a scaling scale.
In one possible implementation, the transformation module is further configured to:
and for each group of image change parameters, rotating the image to be processed according to the rotation angle contained in the image change parameters, and scaling the image to be processed according to the scaling scale contained in the image change parameters to obtain the transformed image to be processed.
In a possible implementation manner, the first feature extraction module is further configured to:
respectively carrying out image feature extraction on the plurality of transformed images to be processed to obtain feature maps corresponding to the plurality of transformed images to be processed;
and aiming at any pixel point, obtaining a first group of feature maps of the pixel point according to the feature vectors of the pixel point in the feature maps corresponding to the transformed images to be processed.
In a possible implementation manner, the second feature extraction module is further configured to:
and performing group equal variation convolution processing on the first group feature graph corresponding to any pixel point to obtain a second group feature graph corresponding to the pixel point, wherein the second group feature graph comprises group structure information of the first group feature graph.
In one possible implementation manner, the second feature extraction module is further configured to:
for each group of image transformation parameters, determining a neighborhood of the image transformation parameter from each group of image transformation parameters;
determining a plurality of convolution regions in the first group of feature maps according to the neighborhood corresponding to each image transformation parameter aiming at any pixel point;
and performing convolution operation on the plurality of convolution areas to obtain a second group feature map corresponding to the pixel point.
In one possible implementation manner, the processing module is further configured to:
and performing pooling processing on the second group feature map to obtain an image descriptor of the image to be processed.
In a possible implementation manner, the second feature extraction module is further configured to:
performing twice group equal variation convolution processing on the first group feature graph corresponding to any pixel point to obtain a first sub-graph and a second sub-graph corresponding to the pixel point, wherein the first sub-graph corresponds to a result of the first group equal variation convolution processing, and the second sub-graph corresponds to a result of the second group equal variation convolution processing;
the processing module is further configured to:
and carrying out bilinear interpolation pooling processing according to the first subgraph and the second subgraph to obtain an image descriptor of the image to be processed.
In one possible implementation, the apparatus is implemented by a neural network.
According to an aspect of the present disclosure, there is provided an electronic device including: a processor; a memory for storing processor-executable instructions; wherein the processor is configured to invoke the memory-stored instructions to perform the above-described method.
According to an aspect of the present disclosure, there is provided a computer readable storage medium having stored thereon computer program instructions which, when executed by a processor, implement the above-described method.
Therefore, the image to be processed can be converted through at least one group of image conversion parameters to obtain a plurality of converted images to be processed, image feature extraction is respectively carried out on the plurality of converted images to be processed to obtain a first group feature map corresponding to each pixel point in the images to be processed, group structure information extraction is carried out on the first group feature map to obtain a second group feature map corresponding to the pixel point, and an image descriptor of the images to be processed can be obtained according to the second group feature map. According to the image processing method and device, the electronic device and the storage medium, the first group of feature maps of each pixel are obtained through a plurality of transformed images to be processed, feature transformation of the images to be processed under a plurality of groups of image transformation parameters is utilized, the information amount is larger, therefore, the accuracy of the image descriptors and the robustness of scale and rotation transformation can be improved, the image descriptors of each pixel point can be obtained through one-time reasoning, and therefore the calculation efficiency of extracting the image descriptors is improved.
It is to be understood that both the foregoing general description and the following detailed description are exemplary and explanatory only and are not restrictive of the disclosure. Other features and aspects of the present disclosure will become apparent from the following detailed description of exemplary embodiments, which proceeds with reference to the accompanying drawings.
Drawings
The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate embodiments consistent with the present disclosure and, together with the description, serve to explain the principles of the disclosure.
FIG. 1 shows a flow diagram of an image processing method according to an embodiment of the present disclosure;
FIG. 2 shows a schematic diagram of an exemplary image processing method according to the present disclosure;
FIG. 3 shows a schematic diagram of an exemplary image processing method according to the present disclosure;
fig. 4 shows a block diagram of an image processing apparatus according to an embodiment of the present disclosure;
FIG. 5 shows a block diagram of an electronic device 800 in accordance with an embodiment of the disclosure;
fig. 6 illustrates a block diagram of an electronic device 1900 in accordance with an embodiment of the disclosure.
Detailed Description
Various exemplary embodiments, features and aspects of the present disclosure will be described in detail below with reference to the accompanying drawings. In the drawings, like reference numbers can indicate functionally identical or similar elements. While the various aspects of the embodiments are presented in drawings, the drawings are not necessarily drawn to scale unless specifically indicated.
The word "exemplary" is used exclusively herein to mean "serving as an example, embodiment, or illustration. Any embodiment described herein as "exemplary" is not necessarily to be construed as preferred or advantageous over other embodiments.
The term "and/or" herein is merely an association describing an associated object, meaning that three relationships may exist, e.g., a and/or B, may mean: a exists alone, A and B exist simultaneously, and B exists alone. In addition, the term "at least one" herein means any one of a plurality or any combination of at least two of a plurality, for example, including at least one of A, B, C, and may mean including any one or more elements selected from the group consisting of A, B and C.
Furthermore, in the following detailed description, numerous specific details are set forth in order to provide a better understanding of the present disclosure. It will be understood by those skilled in the art that the present disclosure may be practiced without some of these specific details. In some instances, methods, means, elements and circuits that are well known to those skilled in the art have not been described in detail so as not to obscure the present disclosure.
In the related technology, an image difference Gaussian pyramid is adopted to detect image key points with scale and direction information in an image, and after an area where the image key points are located is cut out from the image according to the scale and the direction of the image key points, an image descriptor of the area is extracted through a convolutional neural network.
Or, in the related technology, the convolutional neural network can be directly adopted to extract the image descriptor, but the convolutional neural network itself does not have scale and rotation invariance, so that the robustness of scale and rotation transformation for the extracted image descriptor is poor.
Fig. 1 illustrates a flowchart of an image processing method according to an embodiment of the present disclosure, which may be performed by a terminal device or other processing device, wherein the terminal device may be a User Equipment (UE), a mobile device, a User terminal, a cellular phone, a cordless phone, a Personal Digital Assistant (PDA), a handheld device, a computing device, a vehicle-mounted device, a wearable device, or the like. The other processing devices may be servers or cloud servers, etc. In some possible implementations, the image processing method may be implemented by a processor calling computer readable instructions stored in a memory.
As shown in fig. 1, the method may include:
in step S11, the to-be-processed image is transformed according to at least one set of image transformation parameters, so as to obtain a plurality of transformed to-be-processed images.
For example, the image transformation parameters may be parameters for guiding the image to be processed to be transformed, the transformation may be a geometric transformation, such as rotation, scaling, and the like, and the image transformation parameters may include: rotation angle and zoom size. The image to be processed can be rotated and scaled according to the scaling size according to the rotation angle in each group of image change parameters, so as to obtain a plurality of transformed images to be processed.
In a possible implementation manner, the method may further include:
determining unit rotation amount and unit zooming amount;
determining at least one rotation angle according to the unit rotation amount and at least one rotation coefficient, and determining at least one scaling scale according to the unit scaling amount and at least one scaling coefficient;
and combining at least one group of image transformation parameters by using the at least one rotation angle and the at least one scaling scale, wherein each group of image transformation parameters comprises a rotation angle and a scaling scale.
For example, the unit rotation amount may be a preset rotation angle, the rotation coefficient may be a parameter for calculating the unit rotation amount accordingly to obtain a new rotation angle, the unit scaling amount may be a preset scaling scale, and the scaling coefficient may be a parameter for calculating the unit scaling amount accordingly to obtain a new scaling scale.
For example, the rotation coefficient may be a multiple of the unit rotation amount or the rotation coefficient may be an exponent of an exponential power operation on the unit rotation amount, and the scaling coefficient may be a multiple of the unit scaling amount or the scaling coefficient may be an exponent of an exponential power operation on the unit scaling amount.
The method comprises the steps of calculating a unit rotation amount through at least one rotation coefficient to obtain at least one rotation angle, calculating a unit scaling amount through at least one scaling coefficient to obtain at least one scaling scale, and combining the at least one rotation angle and the at least one scaling scale to obtain at least one group of image transformation parameters, wherein each group of image transformation parameters comprises a rotation angle and a scaling scale. In this way, a uniform sampling of the image variation parameters can be achieved.
In a possible implementation manner, the transforming the image to be processed according to at least one group of image transformation parameters to obtain a plurality of transformed images to be processed may include:
and for each group of image change parameters, rotating the image to be processed according to the rotation angle contained in the image change parameters, and scaling the image to be processed according to the scaling scale contained in the image change parameters to obtain the transformed image to be processed.
For example, if the image to be processed is preset to rotate 5 times, the unit rotation amount is 360 °/5 — 72 °, and the unit zoom amount of the image to be processed is preset to 1/2, where the length and the width are respectively reduced to the image to be processed, assuming that the rotation coefficients are: 1. 2, 3, 4, 5, the obtained rotation angles include: 72 °, 144 °, 216 °, 288 °, 360 °, assuming the scaling factors: 0. 1, 2, and 3, the obtained scaling scale includes: 1. 1/2, 1/4, 1/8. After the above rotation angle and scaling are combined, 20 sets of image transformation parameters can be obtained, which are: {72 °, 1}, {72 °, 1/2}, {72 °, 1/4}, {72 °, 1/8}, {144 °, 1}, {144 °, 1/2}, {144 °, 1/4}, {144 °, 1/8}, {216 °, 1}, {216 °, 1/2}, {216 °, 1/4}, {216 °, 1/8}, {288 °, 1}, {288 °, 1/2}, {288 °, 1/4}, {288 °, 1/8}, {360 °, 1}, {360 °, 1/2}, {360 °, 1/4}, {360 °, 1/8 }.
After the image to be processed is subjected to image transformation according to the 20 sets of image transformation parameters, 20 transformed images to be processed can be obtained, and for example, fig. 2 can be referred to.
In step S12, image feature extraction is performed on the multiple transformed images to be processed, so as to obtain a first group feature map corresponding to each pixel point in the images to be processed, where the first group feature map corresponding to any pixel point is used to represent a feature vector of the pixel point corresponding to the multiple transformed images to be processed.
For example, a convolutional neural network for image feature extraction may be pre-trained to respectively perform image feature extraction on a plurality of transformed images to be processed to extract a feature map of each image to be processed, and a first group of feature maps of any pixel in the images to be processed may be obtained according to the feature maps of the plurality of images to be processed.
In a possible implementation manner, the performing image feature extraction on the multiple transformed images to be processed to obtain a first group feature map corresponding to each pixel point in the images to be processed respectively may include:
respectively carrying out image feature extraction on the plurality of transformed images to be processed to obtain feature maps corresponding to the plurality of transformed images to be processed;
and aiming at any pixel point, obtaining a first group of feature maps of the pixel point according to the feature vectors of the pixel point in the feature maps corresponding to the transformed images to be processed.
The image feature extraction can be respectively carried out on a plurality of transformed images to be processed to obtain a plurality of feature maps corresponding to the transformed images to be processed, and the feature map corresponding to any transformed image to be processed comprises feature vectors after pixel points in the images to be processed are transformed. For any pixel point in the image to be processed, obtaining a feature vector of the pixel point in the feature map corresponding to the converted image to be processed, where the feature vectors of the pixel point in the feature maps corresponding to all the converted images to be processed form a first group feature map of the pixel point, for example, as shown in fig. 2, fig. 2 schematically shows the first group feature map corresponding to the center point of the image to be processed, where each square in the rightmost side of fig. 2 schematically shows one feature vector corresponding to the center point, and each feature vector corresponds to one group of image transformation parameters.
In step S13, group structure information is extracted from the first group feature map corresponding to any pixel point, and a second group feature map corresponding to the pixel point is obtained.
In step S14, an image descriptor of the to-be-processed image is obtained according to the second group feature map corresponding to each pixel point.
For example, for a first group feature map corresponding to any pixel point, group structure information in the first group feature map may be extracted, and the group structure information may be used to characterize a transformation mode of the pixel point on rotation and scale. The cluster structure information may reflect the change attribute or change mode of the change such as rotation, scaling, etc. experienced by the pixel. Illustratively, the group structure information of the first group feature map may be extracted by a pre-trained convolutional neural network for extracting group structure information to obtain a second group feature map.
After the second group feature map corresponding to each pixel point is obtained, group convolution processing may be performed on the second group feature map to obtain an image descriptor of each pixel.
Therefore, the image to be processed can be converted through at least one group of image conversion parameters to obtain a plurality of converted images to be processed, image feature extraction is respectively carried out on the plurality of converted images to be processed to obtain a first group feature map corresponding to each pixel point in the images to be processed, group structure information extraction is carried out on the first group feature map to obtain a second group feature map corresponding to the pixel point, and an image descriptor of the images to be processed can be obtained according to the second group feature map. According to the image processing method provided by the disclosure, the first group of feature maps of each pixel are obtained through a plurality of transformed images to be processed, the feature transformation of the images to be processed under a plurality of groups of image transformation parameters is utilized, the information quantity is larger, the accuracy of the image descriptors and the robustness of scale and rotation transformation can be improved, the image descriptors of each pixel point can be obtained through one-time reasoning, and the calculation efficiency of extracting the image descriptors is improved.
In a possible implementation manner, the extracting group structure information of the first group feature map corresponding to any pixel point to obtain a second group feature map corresponding to the pixel point includes:
and performing group equal variation convolution processing on the first group feature graph corresponding to any pixel point to obtain a second group feature graph corresponding to the pixel point, wherein the second group feature graph comprises group structure information of the first group feature graph.
For example, a group invariant convolutional neural network for performing a group invariant convolutional process on a first group feature map may be pre-trained, and the first group feature map is processed by the group invariant convolutional neural network to obtain a second group feature map, where the second group feature map may be a matrix with a scale s × r × n, where s is the number of scaling coefficients, r is the number of rotation coefficients, and n is a feature dimension.
Therefore, by carrying out group convolution processing on the first group characteristic diagram, the learned image descriptor can resist geometric transformation caused by visual angle change, the robustness is improved, and the accuracy of the image descriptor can be improved.
In a possible implementation manner, the performing a group invariant convolution process on the first group feature map corresponding to any pixel point to obtain a second group feature map corresponding to the pixel point may include:
for each group of image transformation parameters, determining a neighborhood of the image transformation parameter from each group of image transformation parameters;
determining a plurality of convolution regions in the first group of feature maps according to the neighborhood corresponding to each image transformation parameter aiming at any pixel point;
and performing convolution operation on the plurality of convolution areas to obtain a second group feature map corresponding to the pixel point.
For example, a neighborhood of each set of image transformation parameters may be determined. For any set of image transformation parameters, the sets of image transformation parameters that are closer to the image transformation parameter may constitute a neighborhood of the set of image transformation parameters, for example: the image transformation parameter 1 obtained by rotating the unit rotation amount clockwise while the scale of the set of image transformation parameters is unchanged, the image transformation parameter 2 obtained by rotating the unit rotation amount counterclockwise while the scale of the set of image transformation parameters is unchanged, the image transformation parameter 3 obtained by reducing the unit scaling amount without rotating the set of image transformation parameters, the image transformation parameter 4 obtained by enlarging the unit scaling amount without rotating the set of image transformation parameters, the image transformation parameter 5 obtained by rotating the unit rotation amount clockwise and enlarging the unit scaling amount scale while the set of image transformation parameters is rotated, the image transformation parameter 6 obtained by rotating the unit rotation amount clockwise and reducing the unit scaling amount scale while the set of image transformation parameters is rotated counterclockwise and the unit scaling amount scale up, the image transformation parameter 7 obtained by rotating the unit rotation amount counterclockwise and reducing the unit scaling amount scale while the set of image transformation parameters are determined, and the image transformation parameter obtained by rotating the unit rotation amount counterclockwise and reducing the unit scaling amount scale down while the set of image transformation parameters is determined The number 8 constitutes a neighborhood of the set of image transformation parameters.
For a first group feature map corresponding to any pixel point, each feature vector in the first group feature map corresponds to one image transformation parameter, a region formed by feature vectors corresponding to a plurality of image transformation parameters in a neighborhood in the first group feature map is determined to be a convolution region, a plurality of neighborhoods can be obtained according to a plurality of groups of image transformation parameters, and a plurality of convolution regions can be determined in the first group feature map according to the plurality of neighborhoods. And respectively carrying out convolution operation on the convolution areas by adopting the same convolution network, and merging the obtained results of the convolution operations to obtain a second group characteristic diagram corresponding to the pixel point.
For example, for a certain position g in the first group of feature maps, the convolved features can be expressed as formula one:
wherein H represents an image transformation parameter, H represents a set of sets of image transformation parameters,
showing the resulting characteristic diagram of layer l-1,
is located at
A feature vector obtained by transforming a position g on the feature map by h, i represents the dimension of the feature vector, and wi(h) Is a weight, [ fl(g)]iRepresenting the feature vector obtained after convolution of the position g, biAre constants obtained by learning.
In a possible implementation manner, the obtaining the image descriptor of the image to be processed according to the second group feature map may include:
and performing pooling processing on the second group feature map to obtain an image descriptor of the image to be processed.
For example, after obtaining the second group feature map, the second group feature map may be a matrix with a size s x r x n. After pooling the matrix (for example, maximum pooling or average pooling may be adopted), the pooled matrix is expanded, and the obtained feature vector is the image descriptor of the pixel.
Therefore, the first group feature map is subjected to group convolution processing in the mode, the learned image descriptor can resist geometric transformation caused by visual angle change, robustness is improved, and accuracy of the image descriptor can be improved.
In a possible implementation manner, the extracting group structure information of the first group feature map corresponding to any pixel point to obtain a second group feature map corresponding to the pixel point may include:
performing twice group equal variation convolution processing on the first group feature graph corresponding to any pixel point to obtain a first sub-graph and a second sub-graph corresponding to the pixel point, wherein the first sub-graph corresponds to a result of the first group equal variation convolution processing, and the second sub-graph corresponds to a result of the second group equal variation convolution processing;
the obtaining the image descriptor of the image to be processed according to the second group feature map corresponding to each pixel point may include:
and carrying out bilinear interpolation pooling processing according to the first subgraph and the second subgraph to obtain an image descriptor of the image to be processed.
For example, a first group feature map of two groups of equal-variation convolution neural networks may be pre-trained for group equal-variation convolution processing, such as: and performing group equal variation convolution processing on the first group characteristic graph through a first group equal variation convolution neural network and a second group equal variation convolution neural network respectively to obtain a first subgraph and a second subgraph.
For example, a bilinear interpolation pooling scheme may be used to pool the first sub-graph and the second sub-graph. Illustratively, the first resulting subgraph is: s r n1, the second resulting sub-graph is: s r n2 matrix. And transposing the first sub-image and the second sub-image, multiplying the transposed first sub-image and the transposed second sub-image to obtain a matrix of n1 n2, and expanding the matrix to obtain a feature vector which is determined as an image descriptor of the image to be processed.
In this way, the second group feature map is subjected to pooling processing in a bilinear interpolation pooling mode, and the image descriptor with stronger expression capability can be obtained.
In one possible implementation, the above method may be implemented by a neural network.
For example, a neural network may be pre-trained, where the neural network may include a first convolutional neural network, a second convolutional neural network, and a pooling network, where the first convolutional neural network may be configured to perform image feature extraction on a plurality of transformed images to be processed to obtain a first group feature map corresponding to each pixel (the specific process may refer to the foregoing embodiment, and is not described herein again), the second convolutional neural network may perform structure information extraction on the first group feature map to obtain a second group feature map corresponding to each pixel (the specific process may refer to the foregoing embodiment, and is not described herein again in this disclosure), and the pooling network may perform pooling processing on the second group feature map to obtain an image descriptor of the images to be processed.
In order to enable those skilled in the art to better understand the embodiments of the present disclosure, the present disclosure is illustrated below by way of example in fig. 3.
And transforming the image to be processed by multiple groups of image transformation parameters at different rotation angles and scaling scales to obtain multiple transformed images to be processed. The method comprises the steps of extracting image features of a plurality of transformed images to be processed through a first convolutional neural network to obtain feature maps of the transformed images to be processed, and obtaining a first group of feature maps corresponding to pixel points according to the feature maps of the transformed images to be processed. The first group characteristic graph corresponding to any pixel point is subjected to group equal-variation convolution processing through the first group equal-variation convolution neural network and the second group equal-variation convolution neural network respectively to obtain a second group characteristic graph (a first subgraph and a second subgraph) containing group structure information, and the image descriptor corresponding to the pixel point can be obtained through bilinear pooling of the second group characteristic graph.
It is understood that the above-mentioned method embodiments of the present disclosure can be combined with each other to form a combined embodiment without departing from the logic of the principle, which is limited by the space, and the detailed description of the present disclosure is omitted. Those skilled in the art will appreciate that in the above methods of the specific embodiments, the specific order of execution of the steps should be determined by their function and possibly their inherent logic.
In addition, the present disclosure also provides an image processing apparatus, an electronic device, a computer-readable storage medium, and a program, which can be used to implement any one of the image processing methods provided by the present disclosure, and the descriptions and corresponding descriptions of the corresponding technical solutions and the corresponding descriptions in the methods section are omitted for brevity.
Fig. 4 illustrates a block diagram of an image processing apparatus according to an embodiment of the present disclosure, which includes, as illustrated in fig. 4:
the transformation module 401 may be configured to transform the image to be processed according to at least one group of image transformation parameters to obtain a plurality of transformed images to be processed;
a first feature extraction module 402, configured to perform image feature extraction on the multiple transformed images to be processed respectively, to obtain a first group feature map corresponding to each pixel point in the images to be processed, where the first group feature map corresponding to any pixel point is used to represent a feature vector of the pixel point corresponding to the multiple transformed images to be processed;
the second feature extraction module 403 may be configured to perform group structure information extraction on the first group feature map corresponding to any pixel point, so as to obtain a second group feature map corresponding to the pixel point;
the processing module 404 may be configured to obtain an image descriptor of the image to be processed according to the second group feature map corresponding to each pixel point.
Therefore, the image to be processed can be converted through at least one group of image conversion parameters to obtain a plurality of converted images to be processed, image feature extraction is respectively carried out on the plurality of converted images to be processed to obtain a first group feature map corresponding to each pixel point in the images to be processed, group structure information extraction is carried out on the first group feature map to obtain a second group feature map corresponding to the pixel point, and an image descriptor of the images to be processed can be obtained according to the second group feature map. According to the image processing device provided by the disclosure, the first group of feature maps of each pixel are obtained through a plurality of converted images to be processed, the feature transformation of the images to be processed under a plurality of groups of image transformation parameters is utilized, the information quantity is larger, the accuracy of the image descriptors and the robustness of scale and rotation transformation can be improved, the image descriptors of each pixel point can be obtained through one-time reasoning, and the calculation efficiency of extracting the image descriptors is improved.
In one possible implementation, the apparatus may further include:
the first determining module can be used for determining a unit rotation amount and a unit zooming amount;
a second determining module, configured to determine at least one rotation angle according to the unit rotation amount and at least one rotation coefficient, and determine at least one scaling dimension according to the unit scaling amount and at least one scaling coefficient;
the combining module may be configured to combine at least one set of image transformation parameters with the at least one rotation angle and the at least one scaling scale, where each set of image transformation parameters includes one rotation angle and one scaling scale.
In a possible implementation manner, the transformation module may be further configured to:
and for each group of image change parameters, rotating the image to be processed according to the rotation angle contained in the image change parameters, and scaling the image to be processed according to the scaling scale contained in the image change parameters to obtain the transformed image to be processed.
In a possible implementation manner, the first feature extraction module may be further configured to:
respectively carrying out image feature extraction on the plurality of transformed images to be processed to obtain feature maps corresponding to the plurality of transformed images to be processed;
and aiming at any pixel point, obtaining a first group of feature maps of the pixel point according to the feature vectors of the pixel point in the feature maps corresponding to the transformed images to be processed.
In a possible implementation manner, the second feature extraction module may be further configured to:
and performing group equal variation convolution processing on the first group feature graph corresponding to any pixel point to obtain a second group feature graph corresponding to the pixel point, wherein the second group feature graph comprises group structure information of the first group feature graph.
In one possible implementation manner, the second feature extraction module may be further configured to:
for each group of image transformation parameters, determining a neighborhood of the image transformation parameter from each group of image transformation parameters;
determining a plurality of convolution regions in the first group of feature maps according to the neighborhood corresponding to each image transformation parameter aiming at any pixel point;
and performing convolution operation on the plurality of convolution areas to obtain a second group feature map corresponding to the pixel point.
In a possible implementation manner, the processing module may be further configured to:
and performing pooling processing on the second group feature map to obtain an image descriptor of the image to be processed.
In a possible implementation manner, the second feature extraction module may be further configured to:
performing twice group equal variation convolution processing on the first group feature graph corresponding to any pixel point to obtain a first sub-graph and a second sub-graph corresponding to the pixel point, wherein the first sub-graph corresponds to a result of the first group equal variation convolution processing, and the second sub-graph corresponds to a result of the second group equal variation convolution processing;
the processing module is further configured to:
and carrying out bilinear interpolation pooling processing according to the first subgraph and the second subgraph to obtain an image descriptor of the image to be processed.
In one possible implementation, the apparatus may be implemented by a neural network.
In some embodiments, functions of or modules included in the apparatus provided in the embodiments of the present disclosure may be used to execute the method described in the above method embodiments, and specific implementation thereof may refer to the description of the above method embodiments, and for brevity, will not be described again here.
Embodiments of the present disclosure also provide a computer-readable storage medium having stored thereon computer program instructions, which when executed by a processor, implement the above-mentioned method. The computer readable storage medium may be a non-volatile computer readable storage medium.
An embodiment of the present disclosure further provides an electronic device, including: a processor; a memory for storing processor-executable instructions; wherein the processor is configured to invoke the memory-stored instructions to perform the above-described method.
The embodiments of the present disclosure also provide a computer program product, which includes computer readable code, and when the computer readable code runs on a device, a processor in the device executes instructions for implementing the picture search method provided in any of the above embodiments.
The embodiments of the present disclosure also provide another computer program product for storing computer readable instructions, which when executed, cause a computer to perform the operations of the picture searching method provided in any of the above embodiments.
The electronic device may be provided as a terminal, server, or other form of device.
Fig. 5 illustrates a block diagram of an electronic device 800 in accordance with an embodiment of the disclosure. For example, the electronic device 800 may be a mobile phone, a computer, a digital broadcast terminal, a messaging device, a game console, a tablet device, a medical device, a fitness device, a personal digital assistant, or the like terminal.
Referring to fig. 5, electronic device 800 may include one or more of the following components: processing component 802, memory 804, power component 806, multimedia component 808, audio component 810, input/output (I/O) interface 812, sensor component 814, and communication component 816.
The processing component 802 generally controls overall operation of the electronic device 800, such as operations associated with display, telephone calls, data communications, camera operations, and recording operations. The processing components 802 may include one or more processors 820 to execute instructions to perform all or a portion of the steps of the methods described above. Further, the processing component 802 can include one or more modules that facilitate interaction between the processing component 802 and other components. For example, the processing component 802 can include a multimedia module to facilitate interaction between the multimedia component 808 and the processing component 802.
The memory 804 is configured to store various types of data to support operations at the electronic device 800. Examples of such data include instructions for any application or method operating on the electronic device 800, contact data, phonebook data, messages, pictures, videos, and so forth. The memory 804 may be implemented by any type or combination of volatile or non-volatile memory devices such as Static Random Access Memory (SRAM), electrically erasable programmable read-only memory (EEPROM), erasable programmable read-only memory (EPROM), programmable read-only memory (PROM), read-only memory (ROM), magnetic memory, flash memory, magnetic or optical disks.
The power supply component 806 provides power to the various components of the electronic device 800. The power components 806 may include a power management system, one or more power supplies, and other components associated with generating, managing, and distributing power for the electronic device 800.
The multimedia component 808 includes a screen that provides an output interface between the electronic device 800 and a user. In some embodiments, the screen may include a Liquid Crystal Display (LCD) and a Touch Panel (TP). If the screen includes a touch panel, the screen may be implemented as a touch screen to receive an input signal from a user. The touch panel includes one or more touch sensors to sense touch, slide, and gestures on the touch panel. The touch sensor may not only sense the boundary of a touch or slide action, but also detect the duration and pressure associated with the touch or slide operation. In some embodiments, the multimedia component 808 includes a front facing camera and/or a rear facing camera. The front camera and/or the rear camera may receive external multimedia data when the electronic device 800 is in an operation mode, such as a shooting mode or a video mode. Each front camera and rear camera may be a fixed optical lens system or have a focal length and optical zoom capability.
The audio component 810 is configured to output and/or input audio signals. For example, the audio component 810 includes a Microphone (MIC) configured to receive external audio signals when the electronic device 800 is in an operational mode, such as a call mode, a recording mode, and a voice recognition mode. The received audio signals may further be stored in the memory 804 or transmitted via the communication component 816. In some embodiments, audio component 810 also includes a speaker for outputting audio signals.
The I/O interface 812 provides an interface between the processing component 802 and peripheral interface modules, which may be keyboards, click wheels, buttons, etc. These buttons may include, but are not limited to: a home button, a volume button, a start button, and a lock button.
The sensor assembly 814 includes one or more sensors for providing various aspects of state assessment for the electronic device 800. For example, the sensor assembly 814 may detect an open/closed state of the electronic device 800, the relative positioning of components, such as a display and keypad of the electronic device 800, the sensor assembly 814 may also detect a change in the position of the electronic device 800 or a component of the electronic device 800, the presence or absence of user contact with the electronic device 800, orientation or acceleration/deceleration of the electronic device 800, and a change in the temperature of the electronic device 800. Sensor assembly 814 may include a proximity sensor configured to detect the presence of a nearby object without any physical contact. The sensor assembly 814 may also include a light sensor, such as a CMOS or CCD image sensor, for use in imaging applications. In some embodiments, the sensor assembly 814 may also include an acceleration sensor, a gyroscope sensor, a magnetic sensor, a pressure sensor, or a temperature sensor.
The communication component 816 is configured to facilitate wired or wireless communication between the electronic device 800 and other devices. The electronic device 800 may access a wireless network based on a communication standard, such as WiFi, 2G or 3G, or a combination thereof. In an exemplary embodiment, the communication component 816 receives a broadcast signal or broadcast related information from an external broadcast management system via a broadcast channel. In an exemplary embodiment, the communication component 816 further includes a Near Field Communication (NFC) module to facilitate short-range communications. For example, the NFC module may be implemented based on Radio Frequency Identification (RFID) technology, infrared data association (IrDA) technology, Ultra Wideband (UWB) technology, Bluetooth (BT) technology, and other technologies.
In an exemplary embodiment, the electronic device 800 may be implemented by one or more Application Specific Integrated Circuits (ASICs), Digital Signal Processors (DSPs), Digital Signal Processing Devices (DSPDs), Programmable Logic Devices (PLDs), Field Programmable Gate Arrays (FPGAs), controllers, micro-controllers, microprocessors or other electronic components for performing the above-described methods.
In an exemplary embodiment, a non-transitory computer-readable storage medium, such as the memory 804, is also provided that includes computer program instructions executable by the processor 820 of the electronic device 800 to perform the above-described methods.
Fig. 6 illustrates a block diagram of an electronic device 1900 in accordance with an embodiment of the disclosure. For example, the electronic device 1900 may be provided as a server. Referring to fig. 6, electronic device 1900 includes a processing component 1922 further including one or more processors and memory resources, represented by memory 1932, for storing instructions, e.g., applications, executable by processing component 1922. The application programs stored in memory 1932 may include one or more modules that each correspond to a set of instructions. Further, the processing component 1922 is configured to execute instructions to perform the above-described method.
The electronic device 1900 may also include a power component 1926 configured to perform power management of the electronic device 1900, a wired or wireless network interface 1950 configured to connect the electronic device 1900 to a network, and an input/output (I/O) interface 1958. The electronic device 1900 may operate based on an operating system stored in memory 1932, such as Windows Server, Mac OS XTM, UnixTM, LinuxTM, FreeBSDTM, or the like.
In an exemplary embodiment, a non-transitory computer readable storage medium, such as the memory 1932, is also provided that includes computer program instructions executable by the processing component 1922 of the electronic device 1900 to perform the above-described methods.
The present disclosure may be systems, methods, and/or computer program products. The computer program product may include a computer-readable storage medium having computer-readable program instructions embodied thereon for causing a processor to implement various aspects of the present disclosure.
The computer readable storage medium may be a tangible device that can hold and store the instructions for use by the instruction execution device. The computer readable storage medium may be, for example, but not limited to, an electronic memory device, a magnetic memory device, an optical memory device, an electromagnetic memory device, a semiconductor memory device, or any suitable combination of the foregoing. More specific examples (a non-exhaustive list) of the computer readable storage medium would include the following: a portable computer diskette, a hard disk, a Random Access Memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), a Static Random Access Memory (SRAM), a portable compact disc read-only memory (CD-ROM), a Digital Versatile Disc (DVD), a memory stick, a floppy disk, a mechanical coding device, such as punch cards or in-groove projection structures having instructions stored thereon, and any suitable combination of the foregoing. Computer-readable storage media as used herein is not to be construed as transitory signals per se, such as radio waves or other freely propagating electromagnetic waves, electromagnetic waves propagating through a waveguide or other transmission medium (e.g., optical pulses through a fiber optic cable), or electrical signals transmitted through electrical wires.
The computer-readable program instructions described herein may be downloaded from a computer-readable storage medium to a respective computing/processing device, or to an external computer or external storage device via a network, such as the internet, a local area network, a wide area network, and/or a wireless network. The network may include copper transmission cables, fiber optic transmission, wireless transmission, routers, firewalls, switches, gateway computers and/or edge servers. The network adapter card or network interface in each computing/processing device receives computer-readable program instructions from the network and forwards the computer-readable program instructions for storage in a computer-readable storage medium in the respective computing/processing device.
The computer program instructions for carrying out operations of the present disclosure may be assembler instructions, Instruction Set Architecture (ISA) instructions, machine-related instructions, microcode, firmware instructions, state setting data, or source or object code written in any combination of one or more programming languages, including an object oriented programming language such as Smalltalk, C + + or the like and conventional procedural programming languages, such as the "C" programming language or similar programming languages. The computer-readable program instructions may execute entirely on the user's computer, partly on the user's computer, as a stand-alone software package, partly on the user's computer and partly on a remote computer or entirely on the remote computer or server. In the case of a remote computer, the remote computer may be connected to the user's computer through any type of network, including a Local Area Network (LAN) or a Wide Area Network (WAN), or the connection may be made to an external computer (for example, through the Internet using an Internet service provider). In some embodiments, the electronic circuitry that can execute the computer-readable program instructions implements aspects of the present disclosure by utilizing the state information of the computer-readable program instructions to personalize the electronic circuitry, such as a programmable logic circuit, a Field Programmable Gate Array (FPGA), or a Programmable Logic Array (PLA).
Various aspects of the present disclosure are described herein with reference to flowchart illustrations and/or block diagrams of methods, apparatus (systems) and computer program products according to embodiments of the disclosure. It will be understood that each block of the flowchart illustrations and/or block diagrams, and combinations of blocks in the flowchart illustrations and/or block diagrams, can be implemented by computer-readable program instructions.
These computer-readable program instructions may be provided to a processor of a general purpose computer, special purpose computer, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, create means for implementing the functions/acts specified in the flowchart and/or block diagram block or blocks. These computer-readable program instructions may also be stored in a computer-readable storage medium that can direct a computer, programmable data processing apparatus, and/or other devices to function in a particular manner, such that the computer-readable medium storing the instructions comprises an article of manufacture including instructions which implement the function/act specified in the flowchart and/or block diagram block or blocks.
The computer readable program instructions may also be loaded onto a computer, other programmable data processing apparatus, or other devices to cause a series of operational steps to be performed on the computer, other programmable apparatus or other devices to produce a computer implemented process such that the instructions which execute on the computer, other programmable apparatus or other devices implement the functions/acts specified in the flowchart and/or block diagram block or blocks.
The flowchart and block diagrams in the figures illustrate the architecture, functionality, and operation of possible implementations of systems, methods and computer program products according to various embodiments of the present disclosure. In this regard, each block in the flowchart or block diagrams may represent a module, segment, or portion of instructions, which comprises one or more executable instructions for implementing the specified logical function(s). In some alternative implementations, the functions noted in the block may occur out of the order noted in the figures. For example, two blocks shown in succession may, in fact, be executed substantially concurrently, or the blocks may sometimes be executed in the reverse order, depending upon the functionality involved. It will also be noted that each block of the block diagrams and/or flowchart illustration, and combinations of blocks in the block diagrams and/or flowchart illustration, can be implemented by special purpose hardware-based systems which perform the specified functions or acts, or combinations of special purpose hardware and computer instructions.
The computer program product may be embodied in hardware, software or a combination thereof. In an alternative embodiment, the computer program product is embodied in a computer storage medium, and in another alternative embodiment, the computer program product is embodied in a Software product, such as a Software Development Kit (SDK), or the like.
Having described embodiments of the present disclosure, the foregoing description is intended to be exemplary, not exhaustive, and not limited to the disclosed embodiments. Many modifications and variations will be apparent to those of ordinary skill in the art without departing from the scope and spirit of the described embodiments. The terminology used herein is chosen in order to best explain the principles of the embodiments, the practical application, or improvements made to the technology in the marketplace, or to enable others of ordinary skill in the art to understand the embodiments disclosed herein.
Claims (10)
1. An image processing method, comprising:
transforming the image to be processed according to at least one group of image transformation parameters to obtain a plurality of transformed images to be processed;
respectively extracting image features of the transformed images to be processed to obtain a first group feature map corresponding to each pixel point in the images to be processed, wherein the first group feature map corresponding to any pixel point is used for representing a feature vector of the pixel point corresponding to the transformed images to be processed;
extracting group structure information of the first group feature map corresponding to any pixel point to obtain a second group feature map corresponding to the pixel point;
and obtaining the image descriptor of the image to be processed according to the second group feature map corresponding to each pixel point.
2. The method of claim 1, further comprising:
determining unit rotation amount and unit zooming amount;
determining at least one rotation angle according to the unit rotation amount and at least one rotation coefficient, and determining at least one scaling scale according to the unit scaling amount and at least one scaling coefficient;
and combining at least one group of image transformation parameters by using the at least one rotation angle and the at least one scaling scale, wherein each group of image transformation parameters comprises a rotation angle and a scaling scale.
3. The method according to claim 1 or 2, wherein the performing image feature extraction on the plurality of transformed images to be processed to obtain a first group feature map corresponding to each pixel point in the images to be processed respectively comprises:
respectively carrying out image feature extraction on the plurality of transformed images to be processed to obtain feature maps corresponding to the plurality of transformed images to be processed;
and aiming at any pixel point, obtaining a first group of feature maps of the pixel point according to the feature vectors of the pixel point in the feature maps corresponding to the transformed images to be processed.
4. The method according to any one of claims 1 to 3, wherein the extracting group structure information of the first group feature map corresponding to any pixel point to obtain a second group feature map corresponding to the pixel point comprises:
and performing group equal variation convolution processing on the first group feature graph corresponding to any pixel point to obtain a second group feature graph corresponding to the pixel point, wherein the second group feature graph comprises group structure information of the first group feature graph.
5. The method according to claim 4, wherein the performing group invariant convolution processing on the first group feature map corresponding to any pixel point to obtain a second group feature map corresponding to the pixel point comprises:
for each group of image transformation parameters, determining a neighborhood of the image transformation parameter from each group of image transformation parameters;
determining a plurality of convolution regions in the first group of feature maps according to the neighborhood corresponding to each image transformation parameter aiming at any pixel point;
and performing convolution operation on the plurality of convolution areas to obtain a second group feature map corresponding to the pixel point.
6. The method according to any one of claims 1 to 5, wherein the extracting group structure information of the first group feature map corresponding to any pixel point to obtain a second group feature map corresponding to the pixel point comprises:
performing twice group equal variation convolution processing on the first group feature graph corresponding to any pixel point to obtain a first sub-graph and a second sub-graph corresponding to the pixel point, wherein the first sub-graph corresponds to a result of the first group equal variation convolution processing, and the second sub-graph corresponds to a result of the second group equal variation convolution processing;
the obtaining the image descriptor of the image to be processed according to the second group feature map corresponding to each pixel point includes:
and carrying out bilinear interpolation pooling processing according to the first subgraph and the second subgraph to obtain an image descriptor of the image to be processed.
7. The method of any one of claims 1 to 6, wherein the method is implemented by a neural network.
8. An image processing apparatus characterized by comprising:
the transformation module is used for transforming the image to be processed according to at least one group of image transformation parameters to obtain a plurality of transformed images to be processed;
the first feature extraction module is used for respectively extracting image features of the plurality of transformed images to be processed to obtain a first group feature map corresponding to each pixel point in the images to be processed, wherein the first group feature map corresponding to any pixel point is used for representing a feature vector of the pixel point corresponding to the plurality of transformed images to be processed;
the second feature extraction module is used for extracting group structure information of the first group feature map corresponding to any pixel point to obtain a second group feature map corresponding to the pixel point;
and the processing module is used for obtaining the image descriptor of the image to be processed according to the second group characteristic graph corresponding to each pixel point.
9. An electronic device, comprising:
a processor;
a memory for storing processor-executable instructions;
wherein the processor is configured to invoke the memory-stored instructions to perform the method of any of claims 1 to 7.
10. A computer readable storage medium having computer program instructions stored thereon, which when executed by a processor implement the method of any one of claims 1 to 7.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201911045867.8A CN112749709A (en) | 2019-10-30 | 2019-10-30 | Image processing method and device, electronic equipment and storage medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201911045867.8A CN112749709A (en) | 2019-10-30 | 2019-10-30 | Image processing method and device, electronic equipment and storage medium |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN112749709A true CN112749709A (en) | 2021-05-04 |
Family
ID=75641762
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201911045867.8A Withdrawn CN112749709A (en) | 2019-10-30 | 2019-10-30 | Image processing method and device, electronic equipment and storage medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN112749709A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115456858A (en) * | 2022-09-16 | 2022-12-09 | 深圳思谋信息科技有限公司 | Image processing method, image processing device, computer equipment and computer readable storage medium |
-
2019
- 2019-10-30 CN CN201911045867.8A patent/CN112749709A/en not_active Withdrawn
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115456858A (en) * | 2022-09-16 | 2022-12-09 | 深圳思谋信息科技有限公司 | Image processing method, image processing device, computer equipment and computer readable storage medium |
| CN115456858B (en) * | 2022-09-16 | 2023-07-18 | 深圳思谋信息科技有限公司 | Image processing method, device, computer equipment and computer readable storage medium |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110348537B (en) | Image processing method and device, electronic equipment and storage medium | |
| CN110287874B (en) | Target tracking method and device, electronic equipment and storage medium | |
| US20210012523A1 (en) | Pose Estimation Method and Device and Storage Medium | |
| CN110378976B (en) | Image processing method and device, electronic equipment and storage medium | |
| CN109816764B (en) | Image generation method and device, electronic equipment and storage medium | |
| CN110503689B (en) | Pose prediction method, model training method and model training device | |
| US20200250495A1 (en) | Anchor determination method and apparatus, electronic device, and storage medium | |
| CN110647834A (en) | Human face and human hand correlation detection method and device, electronic equipment and storage medium | |
| WO2021017358A1 (en) | Pose determination method and apparatus, electronic device, and storage medium | |
| CN110889469A (en) | Image processing method and device, electronic equipment and storage medium | |
| CN109948494B (en) | Image processing method and device, electronic equipment and storage medium | |
| CN111340048B (en) | Image processing method and device, electronic equipment and storage medium | |
| CN111507408A (en) | Image processing method and device, electronic equipment and storage medium | |
| TWI778313B (en) | Method and electronic equipment for image processing and storage medium thereof | |
| CN108900903B (en) | Video processing method and device, electronic equipment and storage medium | |
| CN111563138B (en) | Positioning method and device, electronic equipment and storage medium | |
| CN111242303A (en) | Network training method and device, and image processing method and device | |
| CN111539410A (en) | Character recognition method and device, electronic equipment and storage medium | |
| CN111401230A (en) | Attitude estimation method and apparatus, electronic device, and storage medium | |
| CN110633715B (en) | Image processing method, network training method and device and electronic equipment | |
| CN113283343A (en) | Crowd positioning method and device, electronic equipment and storage medium | |
| CN111311588B (en) | Repositioning method and device, electronic equipment and storage medium | |
| CN109635926B (en) | Attention feature acquisition method and device for neural network and storage medium | |
| CN112749709A (en) | Image processing method and device, electronic equipment and storage medium | |
| CN111354444A (en) | Pathological section image display method and device, electronic equipment and storage medium |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| WW01 | Invention patent application withdrawn after publication |
Application publication date: 20210504 |
|
| WW01 | Invention patent application withdrawn after publication |