CN111325216B - Image local feature description method and device, computer equipment and storage medium - Google Patents
Image local feature description method and device, computer equipment and storage medium Download PDFInfo
- Publication number
- CN111325216B CN111325216B CN201811536085.XA CN201811536085A CN111325216B CN 111325216 B CN111325216 B CN 111325216B CN 201811536085 A CN201811536085 A CN 201811536085A CN 111325216 B CN111325216 B CN 111325216B
- Authority
- CN
- China
- Prior art keywords
- image
- local
- subintervals
- local area
- subinterval
- 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.)
- Active
Links
- 238000000034 method Methods 0.000 title claims abstract description 56
- 239000013598 vector Substances 0.000 claims abstract description 54
- 238000012545 processing Methods 0.000 claims description 21
- 238000013507 mapping Methods 0.000 claims description 17
- 238000001514 detection method Methods 0.000 claims description 15
- 238000005070 sampling Methods 0.000 claims description 15
- 238000004422 calculation algorithm Methods 0.000 claims description 12
- 238000004590 computer program Methods 0.000 claims description 8
- 238000012163 sequencing technique Methods 0.000 claims description 5
- 230000008859 change Effects 0.000 abstract description 30
- 238000005286 illumination Methods 0.000 abstract description 10
- 230000000007 visual effect Effects 0.000 abstract description 10
- 238000004364 calculation method Methods 0.000 abstract description 6
- 238000010586 diagram Methods 0.000 description 10
- 239000011159 matrix material Substances 0.000 description 7
- 230000008569 process Effects 0.000 description 7
- 238000004891 communication Methods 0.000 description 6
- PXFBZOLANLWPMH-UHFFFAOYSA-N 16-Epiaffinine Natural products C1C(C2=CC=CC=C2N2)=C2C(=O)CC2C(=CC)CN(C)C1C2CO PXFBZOLANLWPMH-UHFFFAOYSA-N 0.000 description 3
- 230000009471 action Effects 0.000 description 3
- 238000005516 engineering process Methods 0.000 description 3
- 230000009466 transformation Effects 0.000 description 3
- 230000004931 aggregating effect Effects 0.000 description 2
- 230000008901 benefit Effects 0.000 description 2
- 238000010276 construction Methods 0.000 description 2
- 230000006870 function Effects 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 230000003287 optical effect Effects 0.000 description 2
- 230000009467 reduction Effects 0.000 description 2
- 230000001131 transforming effect Effects 0.000 description 2
- 238000007792 addition Methods 0.000 description 1
- 230000002776 aggregation Effects 0.000 description 1
- 238000004220 aggregation Methods 0.000 description 1
- 238000004458 analytical method Methods 0.000 description 1
- 230000005540 biological transmission Effects 0.000 description 1
- 238000012512 characterization method Methods 0.000 description 1
- 238000006243 chemical reaction Methods 0.000 description 1
- 230000008878 coupling Effects 0.000 description 1
- 238000010168 coupling process Methods 0.000 description 1
- 238000005859 coupling reaction Methods 0.000 description 1
- 230000003247 decreasing effect Effects 0.000 description 1
- 230000001419 dependent effect Effects 0.000 description 1
- 238000013461 design Methods 0.000 description 1
- 239000000835 fiber Substances 0.000 description 1
- 239000000203 mixture Substances 0.000 description 1
- 238000010606 normalization Methods 0.000 description 1
- 230000002093 peripheral effect Effects 0.000 description 1
- 238000013139 quantization Methods 0.000 description 1
- 238000012216 screening Methods 0.000 description 1
- 239000004065 semiconductor Substances 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
- 238000012546 transfer Methods 0.000 description 1
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
- G06V10/44—Local feature extraction by analysis of parts of the pattern, e.g. by detecting edges, contours, loops, corners, strokes or intersections; Connectivity analysis, e.g. of connected components
-
- 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]
Abstract
The embodiment of the invention provides a method, a device, computer equipment and a storage medium for describing local features of an image. Wherein the method comprises the following steps: detecting a local area in an image, and dividing the local area into a plurality of subintervals; establishing a feature vector of each subinterval; and obtaining the feature vector of the local area according to the feature vectors of the plurality of subintervals. Therefore, the image is not required to be described by calculating gradient information of the image, the calculation mode is simple, and the influence caused by internal and external factors such as illumination change, visual angle change and scale change can be resisted.
Description
Technical Field
The present invention relates to the field of image processing technologies, and in particular, to a method and apparatus for describing local features of an image, a computer device, and a storage medium.
Background
The image local feature is widely applied in the fields of computer vision, image matching, target recognition and the like, the basic starting point is to have a reliable image corresponding point set, the establishment of the reliable corresponding relation between image points is usually dependent on an excellent local image feature vector, and the core problem of the local image feature vector is invariance and distinguishability, namely, the attack caused by changes of visual angles, illumination, rotation, shielding, noise, scale and the like can be resisted, so that the related application capacity and efficiency based on images in various environments are improved. Some inventive methods regarding this technology have also emerged in recent years:
(1) And obtaining local feature descriptors of the image to be processed, forming a descriptor set, and performing dimension reduction on each local feature descriptor in the descriptor set according to a dimension reduction matrix to obtain low-dimensional local feature descriptors corresponding to each local feature descriptor. The method can reduce the dimension of the local feature descriptors in the prior art and remove redundant information of the local feature descriptors in the prior art.
(2) The method comprises the steps of obtaining local feature descriptors of an image to be processed, sorting all the local feature descriptors according to importance, selecting a plurality of local feature descriptors for aggregating global feature descriptors from the sorted local feature descriptors according to a cutoff threshold, and aggregating the local feature descriptors by using a Gaussian mixture model to obtain global feature descriptors of the image to be processed. The method can reduce the time complexity in the global feature descriptor aggregation process in the prior art and improve the discrimination and the robustness of the global feature descriptors.
(3) And transforming SIFT (Scale Invariant Feature Transform ) descriptors through a transformation matrix to obtain a 16×8 matrix, transforming the obtained 16×8 matrix into 128-dimensional vectors, and obtaining local feature descriptors, wherein the new descriptors have stronger discrimination capability.
(4) And normalizing the obtained local interest region, dividing local blocks by using a polar coordinate sampling grid, carrying out quantization sampling on the local blocks, mapping the local blocks into a 2-dimensional matrix, extracting 2-dimensional DCT (Discrete Cosine Transform ) frequency domain features, scanning the DCT coefficient matrix according to a zig-zag sequence, rearranging and screening the DCT features, and forming a final local descriptor. The method not only reserves the original spatial information of the local block, but also can tolerate certain deformation, and enhances the robustness of the descriptor.
However, the problems that the prior art fails to solve are as follows:
(1) Affine invariance: the lack of stability in dealing with images at large viewing angle variations is mainly due to the fact that feature detection, such as DOG (gaussian difference operator), is limited, and therefore it is critical how to select a feature detection algorithm with affine invariance.
(2) Rotation invariance: although the existing feature description vector has a certain rotation invariance, but is not completely rotation invariance, for example, SIFT, SURF (Speed Up Robust Features, scale rotation invariance), GLOH (Gradient Location and Orientation Histogram) and the like are determined by using principal directions around the key points, the principal directions are determined by principal peak values of gradient histograms of areas around the extreme points, but when rotation change occurs to an image, deviation exists in the statistical principal directions of the histograms, so that the rotation invariance is not in a complete sense.
(3) Computational complexity: the traditional feature algorithm needs to use gradient information, such as SIFT, to divide local areas by using a spatial relationship, and calculates gradient values through pixel information of images, wherein a data calculation flow exists in the steps, and the follow-up timeliness is relatively complicated.
Disclosure of Invention
The embodiment of the invention provides a method, a device, computer equipment and a storage medium for describing local features of an image, which can describe the image without calculating gradient information of the image, has a simple calculation mode, can resist the influence of internal and external factors such as illumination change, visual angle change, scale change and the like, has rotation invariance, and can greatly improve the accuracy and the efficiency in the applications such as image registration, target identification, classification and the like.
In a first aspect, an embodiment of the present invention provides a method for describing local features of an image, where the method includes: detecting a local area in an image, and dividing the local area into a plurality of subintervals; establishing a feature vector of each subinterval; and obtaining the feature vector of the local area according to the feature vectors of the plurality of subintervals.
In a second aspect, an embodiment of the present invention provides a device for describing local features of an image, where the device includes: the detection processing module is used for detecting a local area in the image and dividing the local area into a plurality of subintervals; the establishing and processing module is used for establishing a characteristic vector of each subinterval; and the acquisition processing module is used for acquiring the characteristic vector of the local area according to the characteristic vectors of the plurality of subintervals.
In a third aspect, an embodiment of the present invention provides a computer apparatus, including: at least one processor, at least one memory and computer program instructions stored in the memory, which when executed by the processor, implement the method as in the first aspect of the embodiments described above.
In a fourth aspect, embodiments of the present invention provide a computer readable storage medium having stored thereon computer program instructions which, when executed by a processor, implement a method as in the first aspect of the embodiments described above.
The method, the device, the computer equipment and the storage medium for describing the local features of the image provided by the embodiment of the invention comprise the work of feature detection and feature description, and in the aspect of feature detection, a Hessian-Affine matching algorithm is adopted to detect the local region in the image, and the local feature region in the image is detected by using the Hessian-Affine in consideration of the fact that the detected region is different in size and shape, so that the local region is normalized into a circular region, and the local region is corresponding to the large visual angle change of the image and has scale invariance. In the aspect of feature description, the brightness information (pixel value) of an image is utilized, a plurality of subintervals are divided in a detected local area according to the brightness, the pixel points in the subintervals are digitally combined by utilizing the pixel values of sampling points in the neighborhood of the subintervals, feature vectors of the pixel points are formed through designing a sort map and an index, and finally the feature vectors of the subintervals are synthesized to form the feature vectors of the local area, so that the image rotation and brightness conversion can be unchanged (namely, the image is changed in the two types, and the feature vectors are unchanged).
Drawings
In order to more clearly illustrate the technical solution of the embodiments of the present invention, the drawings that are needed to be used in the embodiments of the present invention will be briefly described, and other drawings can be obtained according to these drawings without inventive effort for a person skilled in the art.
FIG. 1 shows a flow diagram of a method of describing image local features of one embodiment of the present disclosure;
FIG. 2 illustrates feature regions detected by a Hessian-Affine matching algorithm of an embodiment of the present disclosure;
FIG. 3 illustrates a schematic diagram of the division of a local area by image brightness size according to an embodiment of the present disclosure;
FIG. 4 is a schematic diagram of a pixel x and N neighboring pixels around the pixel x according to an embodiment of the disclosure;
FIG. 5 illustrates a schematic diagram of an alignment position of an embodiment of the present disclosure;
FIG. 6 shows a schematic diagram of an index build table and vector construction of an embodiment of the present disclosure;
FIG. 7 shows a schematic diagram of a new position where point x is rotated in accordance with an embodiment of the present disclosure;
FIG. 8 shows a parameter comparison schematic of the comparison results of an embodiment of the present disclosure;
FIG. 9 illustrates a contrast diagram of image matching under change in viewing angle for an embodiment of the present disclosure;
FIG. 10 illustrates a contrast schematic of image matching under blur in an embodiment of the present disclosure;
FIG. 11 illustrates a graph of image matching contrast under illumination variation in accordance with an embodiment of the present disclosure;
FIG. 12 illustrates a contrast schematic of image matching under scale variation of an embodiment of the present disclosure;
FIG. 13 illustrates an image matching contrast schematic under rotational variation of an embodiment of the present disclosure;
FIG. 14 shows a schematic block diagram of a description apparatus of image local features of one embodiment of the disclosure;
fig. 15 shows a schematic hardware structure of a computer device according to an embodiment of the present invention.
Detailed Description
Features and exemplary embodiments of various aspects of the present invention will be described in detail below, and in order to make the objects, technical solutions and advantages of the present invention more apparent, the present invention will be described in further detail below with reference to the accompanying drawings and examples. It should be understood that the specific embodiments described herein are merely configured to illustrate the invention and are not configured to limit the invention. It will be apparent to one skilled in the art that the present invention may be practiced without some of these specific details. The following description of the embodiments is merely intended to provide a better understanding of the invention by showing examples of the invention.
It is noted that relational terms such as first and second, and the like are used solely to distinguish one entity or action from another entity or action without necessarily requiring or implying any actual such relationship or order between such entities or actions. Moreover, the terms "comprises," "comprising," or any other variation thereof, are intended to cover a non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements does not include only those elements but may include other elements not expressly listed or inherent to such process, method, article, or apparatus. Without further limitation, an element defined by the phrase "comprising … …" does not exclude the presence of other like elements in a process, method, article or apparatus that comprises the element.
An embodiment of the first aspect of the present invention proposes a method for describing local features of an image, and fig. 1 shows a schematic flow chart of a method for describing local features of an image according to an embodiment of the present disclosure. Wherein the method comprises the following steps:
step 102, detecting a local area in an image.
In some embodiments, detecting the local region in the image is followed by: the local area is normalized to a circular area.
The step 102 is a feature detection step, and adopts a Hessian-Affine matching algorithm for detection. The key point areas detected by the method are abundant in quantity, and particularly have good stability and robustness to noise; in order to cope with geometric deformation generated under large visual angle change, an elliptic shape is used for a characteristic region positioned by a Hessian-Affine matching algorithm, and image content in the elliptic shape can be kept unchanged under different visual angles, so that scale and Affine invariance are considered. Fig. 2 shows the feature areas detected by the Hessian-Affine matching algorithm.
And 104, dividing the local area into a plurality of subintervals, and establishing a characteristic vector of each subinterval.
In some embodiments, the dividing the local area into a plurality of subintervals specifically includes: acquiring brightness information of an image; the local area is divided into a plurality of subintervals using the luminance information.
In some embodiments, the establishing a feature vector of each subinterval specifically includes: for each subinterval, a feature vector of the pixel points of the subinterval is established by using the pixel values of the sampling points in the neighborhood subinterval of the subinterval.
In some embodiments, using pixel values of sampling points in a neighborhood subinterval of the subinterval, a feature vector of the pixel points of the subinterval is established, and specifically includes: for the pixel points in each subinterval, constructing a group of digital combinations by using the pixel values of the sampling points in the neighborhood subinterval of the subinterval; establishing a mapping relation, and mapping the digital combination into a sequencing combination according to the mapping relation; and generating the feature vector of the pixel points of the subinterval according to the ordered combination.
This step 104 is a feature description step of normalizing the detected partial image area to a circle and dividing it into B sub-intervals bin according to the brightness size, as shown in fig. 3, taking b=6 as an example, the pixels in each sub-interval are continuous and all within a certain interval.
As shown in fig. 4, for one pixel point X (X 0 As the center of a circle in the local region), a set of digital combinations P are constructed using the pixel values of N sampling points (n=4, for example) in its neighborhood, then the combination is expressed as P (x) = (I (x) 1 ),I(x 2 ),I(x 3 ),I(x 4 ),......,I(x N ) And) wherein I (x i ) Representing the pixel value of the i (i e N) th sample point. Specifically, when n=4, the 4 points are collected to obtain the corresponding pixel value I (x 1 )＝86，I(x 2 )＝217，I(x 3 )＝152，I(x 4 ) =101, then the corresponding combination of points x is p= (86,217,152,101).
Simultaneously constructing a mapping relation, and mapping the digital combination P into a non-descending arrangement lambda (P) =pi (lambda) by a function 1 ,λ 2 ,λ 3 ,λ 4 ,......,λ N ) Wherein lambda is i ＝λ(I(x i ) I e N represents the ordering position of the ith sample point in the permutation. According to the above mapping relation, the combination P of the points x is mapped to be arranged pi, and the pixel value of the point x can be large according to 4 sampling pointsSmall order ranking, as shown in fig. 5, this ranking may be denoted pi (1,4,3,2). The meaning is as follows: i (x) 1 )＜I(x 4 )＜I(x 3 )＜I(x 2 ) According to this non-descending order, subscripts are taken as an arrangement.
For permutation and combination, N (1, 2,3,.. The term "N") should have N ≡! (factorial of N) permutations, an index table is designed. As shown in fig. 6, if the up-sampling point n=4, there are 24 arrangements, then there are 24 dimensions of the vector for the pixel point x, and if pi (1,4,3,2) corresponds to the 6 th bit, then the 6 th bit in the vector is 1, and the rest is 0, then it can be represented as phi (pi (1,4,3,2))= (0,0,0,0,0,1,0,0.
And 106, obtaining the feature vector of the local area according to the feature vectors of the plurality of subintervals.
In some embodiments, obtaining the feature vector of the local area according to the feature vectors of the plurality of subintervals specifically includes: and combining the feature vectors of the plurality of subintervals to generate the feature vector of the local area.
The pixels within each sub-region are represented in the above manner and finally added together as in equation (1), then the 6 sub-regions are combined together for a total of 24 x 6 = 144 dimension eigenvectors, the local region being represented by N-! X b=4-! X 6 = 144 dimensions (B = 6, n = 4, x e bin in the note column) i ，bin i ∈B)，des i Representing a certain point x of the subinterval i Is a feature vector of the local area.
descriptor＝(des 1 ,des 2 ,....des B )
The above has been described as having no distortion in image rotation. Assuming that the image has rotated, the position of point x is rotated to x ', as shown in fig. 7, at which time the digital combination P ' = (I (x ' 1 ),I(x' 2 ),I(x' 3 ),I(x' 4 ) But rotated pictureThe pixel values of the image are also unchanged, so P '= (86,217,152,101), its corresponding arrangement is still n (1,4,3,2), the eigenvector of x' is still (0,0,0,0,0,1,0,0, a.the.the.m., 0) equal to the vector of point x, in the sense that there is complete rotation-invariant for the image, and the rotation-invariant is independent of this eigenvector, unrestricted (SIFT determines one principal direction from the spatial division, estimates the direction of local uniformity of each local block, and constructs the description vector of the relative orientation to achieve rotation invariance). Similarly, if the image is subjected to photometric transformation, the pixel values of each point are correspondingly changed, for example, the image is slightly darkened, and if the image is slightly darkened, all the pixel points on the image are reduced by the same degree, even if nonlinear brightness change occurs, the overall arrangement sequence is unchanged, so that the feature vector is unchanged, and the illumination change invariance is further provided.
The invention uses Hessian-Affine to detect local feature areas in images, which can cope with large visual angle changes and has scale invariance, and in the aspect of feature description, the invention is different from the traditional descriptor that uses space division relation, based on one type of image brightness information, the local area is divided into a plurality of subintervals bin, the pixel points of each subinterval are digitally combined and sequenced and mapped by using the pixel values of sampling points in the neighborhood of the pixel points, and an index catalog is designed based on the pixel points to represent the feature vector of the pixel points, so that the image rotation and brightness transformation can be unchanged (namely, the image is changed in the two types, the feature vector is unchanged), and compared with the traditional image feature vector based on gradient information, the embodiment of the invention does not use data calculation, simplifies the design flow, and has higher timeliness in theory.
Compared with the prior art, the embodiment of the invention realizes the description method of the local image characteristics, has simple method and convenient calculation, and has the technical advantages that:
and (3) feature detection: the Hessian-Affine is adopted to detect the local area in the image, so that the large visual angle change of the image can be dealt with, the scale invariance is ensured, the detection speed is higher, and in addition, the feature quantity is rich, so that the image matching under the shielding condition is facilitated. Table 1 shows an analytical comparison with the relevant usual detection algorithm:
TABLE 1
Characterization: the local area is divided into a plurality of subintervals according to the brightness, and the traditional method is to divide the characteristic area in space, and the dividing method is to divide the characteristic area into a plurality of concentric circles according to the radius with the interest point as the center (the local area dividing method of SIFT) only according to a simple space structure. Compared with the traditional method, the embodiment of the invention divides the local characteristic area according to the brightness order, the continuous brightness distribution is unchanged for any monotonically increasing or decreasing brightness change, the spatial relationship is fused, more complex brightness change can be processed, and the illumination change has good resistance; secondly, a feature vector is constructed by a group of digital combinations in the local area, sequencing mapping and index catalog construction, the complete rotation invariance is ensured, and finally, feature vectors are formed by combining all subintervals.
In order to verify the method provided by the embodiment of the invention, image matching comparison is performed by utilizing a plurality of groups of data and common SIFT features. In terms of parameter selection, according to the comparison result shown in fig. 8, b=8, n=4 and b=6, n=4 achieve better accuracy (accuracy), but in order to ensure the dimension is low, the embodiment of the present invention finally adopts the latter parameter, i.e. b=6, n=4 (i.e. dividing the local area into 6 subintervals, and selecting 4 sampling points) as an example, and this parameter is also taken as an example in the description of the embodiment method described above.
The picture adopted by the embodiment comprises visual angle change, image blurring, illumination change, scale change and rotation change. The matching method carries out rough matching through a BBF (Best-Bin-First) algorithm, a plurality of points to be matched of the matching points are determined, then a homography matrix between two images is estimated through a Ransac (Random Sample Consensus) algorithm and used as geometric constraint, and accurate matching between the characteristic points is completed. The results of the image matching comparison are shown in fig. 9 to 13, where fig. 9 is the image matching comparison under the change of the viewing angle, the left image in fig. 9 is the image matching under the change of the viewing angle in the present embodiment, and the right image in fig. 9 is the image matching under the change of the viewing angle of the SIFT feature; FIG. 10 is a comparison of image matching under blur, wherein the left image in FIG. 10 is the image matching under blur of the present embodiment, and the right image in FIG. 10 is the image matching under blur of SIFT feature; FIG. 11 is a comparison of image matching under illumination change, wherein the left image in FIG. 11 is the image matching under illumination change of the present embodiment, and the right image in FIG. 11 is the image matching under illumination change of SIFT feature; FIG. 12 is a comparison of image matching under scale change, the left image in FIG. 12 is the image matching under scale change of the present embodiment, and the right image in FIG. 12 is the image matching under scale change of SIFT feature; fig. 13 is a comparison of image matching under rotation variation, the left image in fig. 13 is the image matching under rotation variation of the present embodiment, and the right image in fig. 13 is the image matching under rotation variation of SIFT features. The bolded black lines in fig. 9-13 represent a mismatch.
The embodiment of the invention aims to improve the robustness and the distinguishing property of the image local feature description, does not adopt complex mathematical calculation, reduces the dimension of the feature vector, and can be used in the fields of image registration, target identification and classification, visual analysis and the like.
An embodiment of the second aspect of the disclosed embodiment proposes a description device of image local features, and fig. 14 shows a schematic block diagram of a description device 140 of image local features of an embodiment of the disclosed embodiment. The apparatus 140 includes a detection processing module 142, a setup processing module 144, and an acquisition processing module 146:
the detection processing module 142 is configured to detect a local area in the image, and divide the local area into a plurality of subintervals.
The establishing processing module 144 is configured to establish a feature vector of each subinterval.
The obtaining processing module 146 is configured to obtain a feature vector of the local area according to the feature vectors of the plurality of subintervals.
In some embodiments, the detection processing module 142 is specifically configured to: and detecting a local area in the image by adopting a Hessian-Affine matching algorithm.
In some embodiments, the image local feature describing device 140 further includes:
and the normalization module is used for normalizing the local area into a circular area.
In some embodiments, the detection processing module 142 is specifically configured to:
brightness information of an image is acquired.
The local area is divided into a plurality of subintervals using the luminance information.
In some embodiments, the processing module 144 is configured to:
for each subinterval, a feature vector of the pixel points of the subinterval is established by using the pixel values of the sampling points in the neighborhood subinterval of the subinterval.
In one embodiment, the processing module 144 is configured to:
for pixel points within each subinterval, a set of digital combinations is constructed using pixel values of sampling points within a neighborhood subinterval of the subinterval.
And establishing a mapping relation, and mapping the digital combination into a sequencing combination according to the mapping relation.
And generating the feature vector of the pixel points of the subinterval according to the ordered combination.
In one embodiment, the processing module 144 is configured to:
and combining the feature vectors of the plurality of subintervals to generate the feature vector of the local area.
In addition, the method for describing the local features of the image according to the embodiment of the present invention described in connection with fig. 1 may be implemented by a computer device. Fig. 15 shows a schematic hardware structure of a computer device according to an embodiment of the present invention. The computer device may include a processor 152 and a memory 154 storing computer program instructions.
In particular, the processor 152 may comprise a Central Processing Unit (CPU), or an application specific integrated circuit (Application Specific Integrated Circuit, ASIC), or may be configured as one or more integrated circuits that implement embodiments of the present invention.
Memory 154 may include mass storage for data or instructions. By way of example, and not limitation, memory 154 may comprise a Hard Disk Drive (HDD), floppy Disk Drive, flash memory, optical Disk, magneto-optical Disk, magnetic tape, or universal serial bus (Universal Serial Bus, USB) Drive, or a combination of two or more of the foregoing. The memory 154 may include removable or non-removable (or fixed) media, where appropriate. The memory 154 may be internal or external to the data processing apparatus, where appropriate. In a particular embodiment, the memory 154 is a non-volatile solid-state memory. In particular embodiments, memory 154 includes Read Only Memory (ROM). The ROM may be mask programmed ROM, programmable ROM (PROM), erasable PROM (EPROM), electrically Erasable PROM (EEPROM), electrically rewritable ROM (EAROM), or flash memory, or a combination of two or more of these, where appropriate.
The processor 152 reads and executes the computer program instructions stored in the memory 154 to implement the method of describing the local features of the image of any of the above embodiments.
In one example, the computer device may also include a communication interface 156 and a bus 158. As shown in fig. 15, the processor 152, the memory 154, and the communication interface 156 are connected to each other via a bus 410 and perform communication with each other.
Communication interface 156 is primarily used to implement communications between modules, devices, units, and/or apparatuses in embodiments of the invention.
Bus 158 includes hardware, software, or both, coupling components of a computer device to each other. By way of example, and not limitation, the buses may include an Accelerated Graphics Port (AGP) or other graphics bus, an Enhanced Industry Standard Architecture (EISA) bus, a Front Side Bus (FSB), a HyperTransport (HT) interconnect, an Industry Standard Architecture (ISA) bus, an infiniband interconnect, a Low Pin Count (LPC) bus, a memory bus, a micro channel architecture (MCa) bus, a Peripheral Component Interconnect (PCI) bus, a PCI-Express (PCI-X) bus, a Serial Advanced Technology Attachment (SATA) bus, a video electronics standards association local (VLB) bus, or other suitable bus, or a combination of two or more of the above. Bus 158 may include one or more buses, where appropriate. Although embodiments of the invention have been described and illustrated with respect to a particular bus, the invention contemplates any suitable bus or interconnect.
In addition, in combination with the method for describing the local features of the image in the above embodiment, the embodiment of the present invention may be implemented by providing a computer readable storage medium. The computer readable storage medium has stored thereon computer program instructions; the computer program instructions, when executed by a processor, implement a method of describing local features of an image in any of the above embodiments.
It should be understood that the invention is not limited to the particular arrangements and instrumentality described above and shown in the drawings. For the sake of brevity, a detailed description of known methods is omitted here. In the above embodiments, several specific steps are described and shown as examples. However, the method processes of the present invention are not limited to the specific steps described and shown, and those skilled in the art can make various changes, modifications and additions, or change the order between steps, after appreciating the spirit of the present invention.
The functional blocks shown in the above-described structural block diagrams may be implemented in hardware, software, firmware, or a combination thereof. When implemented in hardware, it may be, for example, an electronic circuit, an Application Specific Integrated Circuit (ASIC), suitable firmware, a plug-in, a function card, or the like. When implemented in software, the elements of the invention are the programs or code segments used to perform the required tasks. The program or code segments may be stored in a machine readable medium or transmitted over transmission media or communication links by a data signal carried in a carrier wave. A "machine-readable medium" may include any medium that can store or transfer information. Examples of machine-readable media include electronic circuitry, semiconductor memory devices, ROM, flash memory, erasable ROM (EROM), floppy disks, CD-ROMs, optical disks, hard disks, fiber optic media, radio Frequency (RF) links, and the like. The code segments may be downloaded via computer networks such as the internet, intranets, etc.
It should also be noted that the exemplary embodiments mentioned in this disclosure describe some methods or systems based on a series of steps or devices. However, the present invention is not limited to the order of the above-described steps, that is, the steps may be performed in the order mentioned in the embodiments, or may be performed in a different order from the order in the embodiments, or several steps may be performed simultaneously.
In the foregoing, only the specific embodiments of the present invention are described, and it will be clearly understood by those skilled in the art that, for convenience and brevity of description, the specific working processes of the systems, modules and units described above may refer to the corresponding processes in the foregoing method embodiments, which are not repeated herein. It should be understood that the scope of the present invention is not limited thereto, and any equivalent modifications or substitutions can be easily made by those skilled in the art within the technical scope of the present invention, and they should be included in the scope of the present invention.
Claims (7)
1. A method of describing local features of an image, the method comprising:
detecting a local area in an image, and dividing the local area into a plurality of subintervals;
for pixel points in each subinterval, constructing a group of digital combinations by using pixel values of sampling points in a neighborhood subinterval of the subinterval;
establishing a mapping relation, and mapping the digital combination into a sequencing combination according to the mapping relation;
generating feature vectors of the pixel points of the subintervals according to the sorting combination;
and combining the feature vectors of the subintervals to generate the feature vector of the local area.
2. The method of claim 1, wherein detecting a local region in an image comprises:
and detecting the local area in the image by adopting a Hessian-Affine matching algorithm.
3. The method of claim 1, wherein the dividing the local region into a plurality of subintervals comprises:
acquiring brightness information of the image;
and dividing the local area into a plurality of subintervals by utilizing the brightness information.
4. A method according to any one of claims 1 to 3, further comprising, after the detecting the local region in the image:
normalizing the local region to a circular region.
5. A device for describing local features of an image, the device comprising:
the detection processing module is used for detecting a local area in the image and dividing the local area into a plurality of subintervals;
the method comprises the steps of establishing a processing module, wherein the processing module is used for constructing a group of digital combinations for pixel points in each subinterval by utilizing pixel values of sampling points in a neighborhood subinterval of the subinterval; establishing a mapping relation, and mapping the digital combination into a sequencing combination according to the mapping relation; generating feature vectors of the pixel points of the subintervals according to the sorting combination;
and the acquisition processing module is used for merging the characteristic vectors of the plurality of subintervals to generate the characteristic vector of the local area.
6. A computer device, comprising: at least one processor, at least one memory and computer program instructions stored in the memory, which when executed by the processor, implement the method of any one of claims 1 to 3.
7. A computer readable storage medium having stored thereon computer program instructions, which when executed by a processor, implement the method of any of claims 1 to 3.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811536085.XA CN111325216B (en) | 2018-12-14 | 2018-12-14 | Image local feature description method and device, computer equipment and storage medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811536085.XA CN111325216B (en) | 2018-12-14 | 2018-12-14 | Image local feature description method and device, computer equipment and storage medium |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN111325216A CN111325216A (en) | 2020-06-23 |
| CN111325216B true CN111325216B (en) | 2024-03-22 |
Family
ID=71166944
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201811536085.XA Active CN111325216B (en) | 2018-12-14 | 2018-12-14 | Image local feature description method and device, computer equipment and storage medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN111325216B (en) |
Families Citing this family (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112529857B (en) * | 2020-12-03 | 2022-08-23 | 重庆邮电大学 | Ultrasonic image diagnosis report generation method based on target detection and strategy gradient |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103295014A (en) * | 2013-05-21 | 2013-09-11 | 上海交通大学 | Image local feature description method based on pixel location arrangement column diagrams |
| CN107506795A (en) * | 2017-08-23 | 2017-12-22 | 国家计算机网络与信息安全管理中心 | A kind of local gray level histogram feature towards images match describes sub- method for building up and image matching method |
| CN108256531A (en) * | 2018-01-05 | 2018-07-06 | 上海交通大学 | A kind of sub- building method of local feature description based on image color information and system |
Family Cites Families (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102144245A (en) * | 2008-08-22 | 2011-08-03 | 惠普开发有限公司 | Image analysis method and system |
-
2018
- 2018-12-14 CN CN201811536085.XA patent/CN111325216B/en active Active
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103295014A (en) * | 2013-05-21 | 2013-09-11 | 上海交通大学 | Image local feature description method based on pixel location arrangement column diagrams |
| CN107506795A (en) * | 2017-08-23 | 2017-12-22 | 国家计算机网络与信息安全管理中心 | A kind of local gray level histogram feature towards images match describes sub- method for building up and image matching method |
| CN108256531A (en) * | 2018-01-05 | 2018-07-06 | 上海交通大学 | A kind of sub- building method of local feature description based on image color information and system |
Non-Patent Citations (1)
| Title |
|---|
| 苗权 ; 程光 ; .面向图像匹配的局部灰度直方图特征描述子.小型微型计算机系统.2016,(07),全文. * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN111325216A (en) | 2020-06-23 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Fan et al. | Registration of optical and SAR satellite images by exploring the spatial relationship of the improved SIFT | |
| CN109784250B (en) | Positioning method and device of automatic guide trolley | |
| CN102722731A (en) | Efficient image matching method based on improved scale invariant feature transform (SIFT) algorithm | |
| TW201142718A (en) | Scale space normalization technique for improved feature detection in uniform and non-uniform illumination changes | |
| Wang et al. | Perceptual hashing-based image copy-move forgery detection | |
| Bodnár et al. | Improving barcode detection with combination of simple detectors | |
| JP2013541119A (en) | System and method for improving feature generation in object recognition | |
| CN107610166B (en) | Planar map image registration method based on linear feature region segmentation | |
| CN108269274B (en) | Image registration method based on Fourier transform and Hough transform | |
| CN108960280B (en) | Picture similarity detection method and system | |
| CN107862319B (en) | Heterogeneous high-light optical image matching error eliminating method based on neighborhood voting | |
| CN108229500A (en) | A kind of SIFT Mismatching point scalping methods based on Function Fitting | |
| CN109472770B (en) | Method for quickly matching image characteristic points in printed circuit board detection | |
| CN108537832B (en) | Image registration method and image processing system based on local invariant gray feature | |
| CN116433733A (en) | Registration method and device between optical image and infrared image of circuit board | |
| CN111325216B (en) | Image local feature description method and device, computer equipment and storage medium | |
| CN111311593A (en) | Multi-ellipse detection and evaluation algorithm, device, terminal and readable storage medium based on image gradient information | |
| CN108960246B (en) | Binarization processing device and method for image recognition | |
| Nawaz et al. | Image authenticity detection using DWT and circular block-based LTrP features | |
| CN113221696A (en) | Image recognition method, system, equipment and storage medium | |
| Kunaver et al. | Image feature extraction-an overview | |
| Temel et al. | ReSIFT: Reliability-weighted sift-based image quality assessment | |
| CN115578594A (en) | Edge positioning method and device based on computer vision and related equipment | |
| Lv et al. | Robust registration of multispectral satellite images based on structural and geometrical similarity | |
| Ren et al. | SAR image matching method based on improved SIFT for navigation system |
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 | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |