CN107154024A - Dimension self-adaption method for tracking target based on depth characteristic core correlation filter - Google Patents

Dimension self-adaption method for tracking target based on depth characteristic core correlation filter Download PDF

Info

Publication number
CN107154024A
CN107154024A CN201710355456.3A CN201710355456A CN107154024A CN 107154024 A CN107154024 A CN 107154024A CN 201710355456 A CN201710355456 A CN 201710355456A CN 107154024 A CN107154024 A CN 107154024A
Authority
CN
China
Prior art keywords
mrow
target
msup
scale
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.)
Pending
Application number
CN201710355456.3A
Other languages
Chinese (zh)
Inventor
刘忠耿
练智超
濮柯佳
李杨
张伟
李敏
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Nanjing University of Science and Technology
Original Assignee
Nanjing University of Science and Technology
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by Nanjing University of Science and Technology filed Critical Nanjing University of Science and Technology
Priority to CN201710355456.3A priority Critical patent/CN107154024A/en
Publication of CN107154024A publication Critical patent/CN107154024A/en
Pending legal-status Critical Current

Links

Classifications

    • 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
    • G06T3/4084—Scaling of whole images or parts thereof, e.g. expanding or contracting in the transform domain, e.g. fast Fourier transform [FFT] domain scaling
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N3/00—Computing arrangements based on biological models
    • G06N3/02—Neural networks
    • G06N3/08—Learning methods
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T7/00—Image analysis
    • G06T7/20—Analysis of motion
    • G06T7/246—Analysis of motion using feature-based methods, e.g. the tracking of corners or segments
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T7/00—Image analysis
    • G06T7/20—Analysis of motion
    • G06T7/269—Analysis of motion using gradient-based methods
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T7/00—Image analysis
    • G06T7/40—Analysis of texture
    • G06T7/41—Analysis of texture based on statistical description of texture
    • G06T7/44—Analysis of texture based on statistical description of texture using image operators, e.g. filters, edge density metrics or local histograms
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T2207/00—Indexing scheme for image analysis or image enhancement
    • G06T2207/20—Special algorithmic details
    • G06T2207/20016—Hierarchical, coarse-to-fine, multiscale or multiresolution image processing; Pyramid transform
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T2207/00—Indexing scheme for image analysis or image enhancement
    • G06T2207/20—Special algorithmic details
    • G06T2207/20081—Training; Learning

Landscapes

  • Engineering & Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • Theoretical Computer Science (AREA)
  • General Physics & Mathematics (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Multimedia (AREA)
  • Biomedical Technology (AREA)
  • Data Mining & Analysis (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Artificial Intelligence (AREA)
  • Probability & Statistics with Applications (AREA)
  • Biophysics (AREA)
  • Computational Linguistics (AREA)
  • Health & Medical Sciences (AREA)
  • Evolutionary Computation (AREA)
  • General Health & Medical Sciences (AREA)
  • Molecular Biology (AREA)
  • Computing Systems (AREA)
  • General Engineering & Computer Science (AREA)
  • Mathematical Physics (AREA)
  • Software Systems (AREA)
  • Image Analysis (AREA)

Abstract

The invention discloses a kind of dimension self-adaption method for tracking target based on depth characteristic core correlation filter.This method comprises the following steps：The convolutional neural networks of pre-training completion are input an image into, depth convolution feature is extracted；Target tracking, using the model of training, estimates position and the yardstick of target；According to the target location of current detection and yardstick, core correlation filter is trained；Using the model update method of adaptive high confidence level, core correlation filter is updated.The present invention is extracted depth convolution feature, using adaptive scale method of estimation and the model modification strategy of adaptive high confidence level, improve the robustness of target following of the target in complex scene and cosmetic variation, target scale change can efficiently and accurately be handled, in addition, due to the model modification strategy using adaptive high confidence level, model following drift has been reduced as far as.

Description

Scale self-adaptive target tracking method based on depth feature kernel correlation filter
Technical Field
The invention relates to the technical field of computer vision, in particular to a scale self-adaptive target tracking method based on a depth feature kernel correlation filter.
Background
In recent years, with the appearance of large-scale labeled data sets and the improvement of computer computing power, deep learning methods, particularly convolutional neural networks, are successfully applied to the computer vision field of image classification, target detection, target recognition, semantic segmentation and the like, which is mainly attributed to the strong target representation capability of the convolutional neural networks. Unlike traditional image features, deep convolution features are learned from a large number of thousands of classes of image data, so high-level convolution features represent semantic features of a target and are applicable to image classification problems. Because the resolution of the convolution features at the high layer is low, the method is not suitable for determining the position of the target, and because of the loss of training data, the training of a depth model in the first few frames of the tracking start is difficult.
Recently, discriminant tracking methods based on correlation filters are of interest to many researchers because of their high efficiency and accuracy in tracking. The tracking method based on the correlation filter trains a correlation filter on line by regressing the input characteristics to the Gaussian distribution of the target, and finds the peak value of the corresponding atlas output by the correlation filter in the subsequent tracking process to determine the position of the target. The related filter skillfully applies fast Fourier transform in operation, so that the calculation complexity is reduced, and the tracking speed is greatly improved. However, the kernel correlation filter algorithm uses the traditional gradient direction histogram feature, and when the appearance representation of the target changes, the tracking algorithm is easy to drift; in addition, the algorithm cannot estimate the scale change of the target, and a simple linear interpolation is adopted to update the model, so that when the target tracking is wrong, the tracking algorithm is shifted due to the final update mechanism.
Disclosure of Invention
The invention aims to provide a scale self-adaptive target tracking method based on a depth feature kernel correlation filter, so that the robustness of target tracking of a target in a complex scene and appearance change is improved, the target scale change is efficiently and accurately processed, and model tracking drift caused by error detection is reduced as much as possible.
The technical solution for realizing the purpose of the invention is as follows: a scale self-adaptive target tracking method based on a depth feature kernel correlation filter comprises the following steps:
step 1, inputting an initial position p of a target0Sum scale s0Setting the window size to be 2.0 times of the target initial bounding box;
step 2, according to the target position p of the t-1 th framet-1Obtaining a target area xt-1The size is the window size;
step 3, extracting a target area xt-1Depth of (2)Convolution characteristic and fast Fourier transform to obtain characteristic spectrumWherein ^ represents a discrete Fourier transform;
step 4, according to the characteristic mapCompute kernel autocorrelation
Step 5, training a position and scale correlation filter;
step 6, according to the position p of the t-1 frame targett-1Obtaining a candidate region z of the target in the t-th frametThe size is the window size;
step 7, extracting candidate region ztAnd performing fast Fourier transform to obtain a feature mapWherein ^ represents a discrete Fourier transform;
step 8, according to the characteristic map of the previous frame of the targetComputing kernel cross-correlation
Step 9, respectively detecting the positions corresponding to the maximum values in the output maps of the position filter and the scale filter to determine the position p of the target in the current frametSum scale st；
And step 10, updating the kernel correlation filter by adopting a self-adaptive model updating strategy.
Further, the method for extracting the deep convolution features in step 3 and step 7 specifically includes the following steps:
(3.1) preprocessing, scaling the window area I to the input size 224 × 224 specified by the convolutional neural network;
(3.2) extracting features, namely extracting feature maps of the 3 rd, 4 th and 5 th convolutional layers of the convolutional neural network;
and (3.3) carrying out bilinear interpolation, and upsampling the extracted 3-layer convolution characteristics to the same size.
Further, the calculating core autocorrelation of step 4And the step 8 of calculating the kernel cross correlationThe specific method comprises the following steps:
(4.1) using a Gaussian kernel, the formula is as follows:
wherein k (x, x ') represents a Gaussian kernel calculated by two feature maps x and x', exp (eta) represents an e index function, sigma is a standard deviation of the Gaussian function, the value of sigma is 0.5, | | |22-normal form representing a vector or matrix;
(4.2) calculating the kernel correlation, the formula is as follows:
wherein k isxx′Representing the nuclear correlation of the characteristic maps x and x', exp (eta) representing an e index function, sigma is the standard deviation of a Gaussian function, the value of sigma is 0.5, | | |. | tory2A 2-normal form representing a vector or matrix,represents the inverse of the discrete fourier transform, denotes the complex conjugate, denotes the discrete fourier transform, and ⊙ denotes the multiplication of the corresponding elements of the two matrices.
Further, the training position and scale correlation filter described in step 5 specifically includes the following steps:
respectively training 1 kernel correlation filter for each layer of feature spectrum according to the deep convolution features extracted in the step 3, and training a model by adopting the following formula:
wherein,representing a feature map convolved according to the l-th layer depthThe obtained correlation filter model is used for calculating the correlation filter model,representation feature mapRepresents discrete Fourier transform, lambda is a regularization parameter to prevent overfitting of the trained model, and lambda is 0.001.
Further, the step 9 of detecting the positions corresponding to the maximum values in the output maps of the position filter and the scale filter, respectively, to determine the position p of the target in the current frametSum scale stThe method comprises the following steps:
(9.1) from the t-th frame image, with position pt-1And selecting candidate area z by window sizet,trans；
(9.2) extracting candidate region zt,trans3 layers depth convolution characteristic map
(9.3) calculating a position filter correlation output confidence coefficient map f for the first layer characteristic mapt,trans (l)：
Wherein f ist (l)The output corresponding map of the position filter representing the characteristic map of the l-th layer,representation feature mapAndthe cross-correlation of the kernels of (a),is the position filter trained and updated from the previous frame,represents an inverse discrete Fourier transform, represents a discrete Fourier transform,representing the multiplication of corresponding elements of two matrixes;
(9.4) outputting the corresponding map ft (3)Starting to estimate the target position from coarse to fine, the target position p of the l-th layert (l)For outputting the corresponding map ft,trans (l)The position corresponding to the maximum value;
(9.5) from the t-th frame image, with position ptSum scale st-1Extracting candidate region z of scale estimatet,sacleConstructing a scale pyramid;
(9.6) extracting gradient direction histogram feature of candidate regionComputing a scale filter correlation output confidence map ft,sacle；
(9.7) target dimension s detected at the t-th frametFor outputting the corresponding map ft,sacleThe maximum value corresponds to the scale.
Further, the updating of the kernel correlation filter by using the adaptive model updating strategy in step 10 is as follows:
(10.1) after completing the estimation of the target position and scale, calculating two tracking result confidence measure criteria, one of which is the peak value of the relevant corresponding output map:
fmax＝maxf
wherein f is a corresponding map output by the correlation between the kernel correlation filter and the candidate region, fmaxIs the peak of the map;
the other is the average peak correlation energy APCE:
wherein f ismaxAnd fminRespectively the maximum and minimum of the output map f, mean () representing the averaging function, fi,jThe ith row and jth column of values representing the output map f;
(10.2) if fmaxAnd APCE are both greater than fmaxAnd APCE historical average value, updating the model, otherwise, not updating; for each depth convolution layer, using a linear interpolation methodUpdating the method, wherein the formula is as follows:
wherein,andthe feature map and the related filter of the previous frame of the ith layer are respectively represented, η is the learning rate, the larger η is, the faster model is updated, and η takes the value of 0.02.
Compared with the prior art, the invention has the following remarkable advantages: (1) the method uses the deep convolution characteristics which are obtained by learning from a large amount of image data of thousands of categories, and has strong target representation capability, so that the algorithm has strong robustness when the target appearance changes and external factors such as illumination change; (2) an adaptive scale estimation method is adopted, the method is similar to position estimation, an independent scale filter is trained, fast Fourier transform is skillfully used, high efficiency is realized, scale estimation is accurate, and the method can be combined into any discriminant tracking algorithm frame; (3) by adopting a self-adaptive model updating strategy with high confidence level, when an error occurs in a target tracking stage, the confidence level of target detection is low, and the model is not updated at the moment, so that the risk of the drift of the tracking algorithm is effectively reduced.
Drawings
FIG. 1 is a flowchart of a scale-adaptive target tracking method based on a depth feature kernel correlation filter according to the present invention.
Fig. 2 is a schematic diagram of extracting deep convolution features.
Fig. 3 is a schematic diagram of target position and scale estimation, wherein (a) is a schematic diagram of target position estimation from coarse to fine, and (b) is a schematic diagram of adaptive target scale estimation.
FIG. 4 is a diagram of adaptive high confidence model update.
FIG. 5 is a plot of the results of the evaluation of the present invention on a standard visual tracking dataset, where (a) is an accuracy plot of the OTB50 dataset, (b) is a correct rate plot of the OTB50 dataset, (c) is an accuracy plot of the OTB100 dataset, and (d) is a correct rate plot of the OTB100 dataset.
Fig. 6 is a graph of the actual video target tracking result of the present invention, where (a) is a graph of the result of Human testing video on the OTB100 data set, (b) is a graph of the result of Walking testing video on the OTB100 data set, (c) is a graph of the result of Tiger testing video on the OTB50 data set, and (d) is a graph of the result of Dog testing video on the OTB50 data set.
Detailed Description
For a better understanding and appreciation of the structural features and advantages achieved by the present invention, reference should be made to the following detailed description of the preferred embodiments taken in conjunction with the accompanying drawings, in which:
with reference to fig. 1, the method for tracking a scale-adaptive target based on a depth feature kernel correlation filter of the present invention includes the following steps:
step 1, inputting an initial position p of a target0Sum scale s0Setting the window size to be 2.0 times of the target initial bounding box;
step 2, according to the target position p of the t-1 th framet-1Obtaining a target area xt-1The size is the window size;
step 3, extracting a target area xt-1Is performed fast, andfast Fourier transform to obtain characteristic spectrumWherein ^ represents a discrete Fourier transform;
step 4, according to the characteristic mapCompute kernel autocorrelation
Step 5, training a position and scale correlation filter;
step 6, according to the position p of the t-1 frame targett-1Obtaining a candidate region z of the target in the t-th frametThe size is the window size;
step 7, extracting candidate region ztAnd performing fast Fourier transform to obtain a feature mapWherein ^ represents a discrete Fourier transform;
step 8, according to the characteristic map of the previous frame of the targetComputing kernel cross-correlation
Step 9, respectively detecting the positions corresponding to the maximum values in the output maps of the position filter and the scale filter to determine the position p of the target in the current frametSum scale st；
And step 10, updating the kernel correlation filter by adopting a self-adaptive model updating strategy.
As a specific example, the method for extracting the deep convolution features in steps 3 and 7 specifically includes the following steps:
(3.1) preprocessing, scaling the window area I to the input size 224 × 224 specified by the convolutional neural network;
(3.2) extracting features, namely extracting feature maps of the 3 rd, 4 th and 5 th convolutional layers of the convolutional neural network;
and (3.3) carrying out bilinear interpolation, and upsampling the extracted 3-layer convolution characteristics to the same size.
As a specific example, the computational core autocorrelation of step 4And the step 8 of calculating the kernel cross correlationThe specific method comprises the following steps:
(4.1) using a Gaussian kernel, the formula is as follows:
wherein k (x, x ') represents a Gaussian kernel calculated by two feature maps x and x', exp (eta) represents an e index function, sigma is a standard deviation of the Gaussian function, the value of sigma is 0.5, | | |22-normal form representing a vector or matrix;
(4.2) calculating the kernel correlation, the formula is as follows:
wherein k isxx′Representing the nuclear correlation of the characteristic maps x and x', exp (eta) representing an e index function, sigma is the standard deviation of a Gaussian function, the value of sigma is 0.5, | | |. | tory2A 2-normal form representing a vector or matrix,represents the inverse of the discrete Fourier transform, represents the complex conjugate, represents the discrete Fourier transform,representing the multiplication of corresponding elements of the two matrices.
As a specific example, the training position and scale correlation filter in step 5 is specifically as follows:
respectively training 1 kernel correlation filter for each layer of feature spectrum according to the deep convolution features extracted in the step 3, and training a model by adopting the following formula:
wherein,representing a feature map convolved according to the l-th layer depthThe obtained correlation filter model is used for calculating the correlation filter model,representation feature mapRepresents discrete Fourier transform, lambda is a regularization parameter to prevent overfitting of the trained model, and lambda is 0.001.
As a specific example, the positions corresponding to the maximum values in the output maps of the position filter and the scale filter are detected respectively in step 9 to determine the position p of the target in the current frametSum scale stThe method comprises the following steps:
(9.1) fromT frame image, with position pt-1And selecting candidate area z by window sizet,trans；
(9.2) extracting candidate region zt,trans3 layers depth convolution characteristic map
(9.3) calculating a position filter correlation output confidence coefficient map f for the first layer characteristic mapt,trans (l)：
Wherein f ist (l)The output corresponding map of the position filter representing the characteristic map of the l-th layer,representation feature mapAndthe cross-correlation of the kernels of (a),is the position filter trained and updated from the previous frame,represents an inverse discrete Fourier transform, represents a discrete Fourier transform,representing the multiplication of corresponding elements of two matrixes;
(9.4) outputting the corresponding map ft (3)Starting to estimate the target position from coarse to fine, the target position p of the l-th layert (l)For outputting the corresponding map ft,trans (l)The position corresponding to the maximum value;
(9.5) from the t-th frame image, with position ptSum scale st-1Extracting candidate region z of scale estimatet,sacleConstructing a scale pyramid;
(9.6) extracting gradient direction histogram feature of candidate regionComputing a scale filter correlation output confidence map ft,sacle；
(9.7) target dimension s detected at the t-th frametFor outputting the corresponding map ft,sacleThe maximum value corresponds to the scale.
As a specific example, the updating the kernel correlation filter by using the adaptive model updating strategy in step 10 is as follows:
(10.1) after completing the estimation of the target position and scale, calculating two tracking result confidence measure criteria, one of which is the peak value of the relevant corresponding output map:
fmax＝maxf
wherein f is a corresponding map output by the correlation between the kernel correlation filter and the candidate region, fmaxIs the peak of the map;
another is the average peak-to-correlation energy (APCE):
wherein f ismaxAnd fminRespectively the maximum and minimum of the output map f, mean () representing the averaging function, fi,jThe ith row and jth column of values representing the output map f;
(10.2) if fmaxAnd APCE are both greater than fmaxAnd APCE historical average value, updating the model, otherwise, not updating; for each depth convolution layer, updating using a linear interpolation method, the formula is as follows:
wherein,andthe feature map and the related filter of the previous frame of the ith layer are respectively represented, η is the learning rate, the larger η is, the faster model is updated, and η takes the value of 0.02.
The invention solves the defect of tracking failure caused by appearance change such as target deformation, illumination change, target rotation and scale change and target shielding. By means of the strong target representation capability of the depth convolution characteristics, the robustness of target tracking of the target in a complex scene and appearance change is improved; in addition, the method can efficiently and accurately process the target scale change, and finally, the model tracking drift caused by error detection is reduced due to the adoption of a self-adaptive model updating strategy with high confidence level.
The present invention will be described in further detail with reference to specific examples.
Example 1
The invention relates to a scale self-adaptive target tracking method based on a depth feature kernel correlation filter, which mainly comprises four steps, wherein the first step is to extract depth convolution features; secondly, training a kernel correlation filter; thirdly, estimating the position and the scale of the target in the current frame; and fourthly, adopting a self-adaptive model updating strategy with high confidence level.
Step 1, inputting an initial position p of a target0Sum scale s0Setting the window size to be 2.0 times of the target initial bounding box;
step 2, according to the target position p of the t-1 th framet-1Obtaining a target area xt-1The size is the window size;
step 3, extracting a target area xt-1The deep convolution characteristic and the fast Fourier transform are carried out to obtain a characteristic mapWherein ^ represents a discrete Fourier transform;
step 4, according to the characteristic mapCompute kernel autocorrelation
Step 5, training a position and scale correlation filter;
step 6, according to the position p of the t-1 frame targett-1Obtaining a candidate region z of the target in the t-th frametThe size is the window size;
step 7, extracting candidate region ztAnd performing fast Fourier transform to obtain a feature mapWherein ^ represents a discrete Fourier transform;
step 8, according to the characteristic map of the previous frame of the targetComputing kernel cross-correlation
Step 9, respectively detecting the positions corresponding to the maximum values of the output corresponding maps of the position filter and the scale filter to determine the position p of the target in the current frametSum scale st；
And step 10, updating the kernel correlation filter by adopting a self-adaptive model updating strategy with high confidence level.
As shown in fig. 2, a schematic diagram of extracting depth convolution features of a scale-adaptive target tracking method based on a depth feature kernel correlation filter is given. The invention relates to a scale self-adaptive target tracking method based on a depth feature kernel correlation filter, which is characterized in that depth convolution features are extracted, are obtained by learning from a large amount of image data of thousands of categories, and have strong target representation capability, so that an algorithm has strong robustness when the appearance of a target changes and external factors such as illumination change, and the specific steps are as follows:
(3.1) preprocessing, scaling the window area I to the input size 224 × 224 specified by the convolutional neural network;
(3.2) extracting features, namely extracting feature maps of the 3 rd, 4 th and 5 th convolutional layers of the convolutional neural network;
and (3.3) carrying out bilinear interpolation, and upsampling the extracted 3-layer convolution characteristics to the same size.
As shown in fig. 3, a schematic diagram of target position and scale estimation of a scale-adaptive target tracking method based on a depth feature kernel correlation filter is given, where fig. 3(a) is a schematic diagram of target position estimation from coarse to fine, and fig. 3(b) is a schematic diagram of adaptive target scale estimation. The invention relates to a scale self-adaptive target tracking method based on a depth characteristic kernel correlation filter, which is characterized by comprising a coarse-to-fine hierarchical position estimation method and a self-adaptive target scale estimation method. The model size of the target in the traditional kernel correlation filtering method is fixed, and the change of the target scale cannot be processed, so that the tracking failure is easily caused. The invention provides a self-adaptive scale estimation method, which has the main idea that an independent scale filter is trained and is estimated through the corresponding scale when the relevant response of the scale filter is maximum, the method skillfully applies fast Fourier transform, is simple and efficient, can be integrated into the traditional discriminant target tracking method, and has the following specific steps:
(9.1) from the t-th frame image, with position pt-1And selecting candidate area z by window sizet,trans；
(9.2) extracting candidate region zt,trans3 layers depth convolution characteristic map
(9.3) calculating a position filter correlation output confidence coefficient map f for the first layer characteristic mapt,trans (l)：
Wherein f ist (l)The output corresponding map of the position filter representing the characteristic map of the l-th layer,representation feature mapAndthe cross-correlation of the kernels of (a),is the position filter trained and updated from the previous frame,represents an inverse discrete Fourier transform, represents a discrete Fourier transform,representing the multiplication of corresponding elements of the two matrices.
(9.4) outputting the corresponding map ft (3)Starting to estimate the target position from coarse to fine, the target position p of the l-th layert (l)For outputting the corresponding map ft,trans (l)The position corresponding to the maximum value;
(9.5) from the t-th frame image, with position ptSum scale st-1Extracting candidate region z of scale estimatet,sacleConstructing a scale pyramid;
(9.6) extracting gradient direction histogram feature of candidate regionComputing a scale filter correlation output confidence map ft,sacle；
(9.7) target dimension s detected at the t-th frametFor outputting the corresponding map ft,sacleThe maximum value corresponds to the scale.
As shown in fig. 4, a schematic diagram of model update with high confidence of self-adaptation of a scale-adaptive target tracking method based on a depth feature kernel correlation filter is provided. The invention relates to a scale self-adaptive target tracking method based on a depth characteristic kernel correlation filter, which is characterized by a self-adaptive high-confidence model updating strategy. The traditional updating method of the nuclear correlation filtering model adopts a simple linear interpolation method, tracking error detection is not carried out, when target detection is wrong during tracking, the model updating can pollute the nuclear correlation filter, and the tracking algorithm drifts. The invention provides two detection confidence degree criteria, when the target tracking is determined to be correct, the model is updated, otherwise, the model is not updated, and the method comprises the following specific steps:
(10.1) after completing the estimation of the target position and scale, calculating two tracking result confidence measure criteria, one of which is the peak value of the relevant corresponding output map:
fmax＝maxf
wherein f is a corresponding map output by the correlation between the kernel correlation filter and the candidate region, fmaxIs the peak of the map. Another is the average peak-to-correlation energy (APCE):
wherein f ismaxAnd fminIs the maximum and minimum of the output map f, mean (or.) represents the averaging function. f. ofi,jThe ith row and jth column values of the output map f are represented.
(10.2) if fmaxAnd APCE are both greater than fmaxAnd APCE historical average value, updating the model, otherwise, not updating. For each depth convolution layer, updating using a linear interpolation method, the formula is as follows:
wherein,andfeature maps and associated filters representing the previous frame of layer l, respectively, η being the learning rate, the larger the model update ηThe faster the ground, η takes on the value 0.02 in the invention.
As shown in fig. 5, a graph of the results of the evaluation of the present invention on standard visual tracking data sets OTB50 and OTB100 is shown, where (a) is an accuracy plot of the OTB50 data set, (b) is a correct rate plot of the OTB50 data set, (c) is an accuracy plot of the OTB100 data set, and (d) is a correct rate plot of the OTB100 data set. The OTB50 dataset has 50 video sequences for a total of 29000 frames, while the OTB100 dataset has 100 video sequences for a total of 58897 frames, each of which has a marker of the target. The evaluation indexes are mainly two types: accuracy and success rate. In the accuracy plots (a) and (c), accuracy is defined as the number of frames between the algorithm detection position and the target calibration position that are no more than 20 pixels from the total evaluation frame number; in the success rate plots (b) and (d), the overlapping rate refers to the percentage of the frame number of the algorithm detection target bounding box and the target calibration bounding box, wherein the percentage of the overlapping part area (intersection operation) between the two bounding boxes accounts for the total area (union operation) and exceeds 50% of the total evaluation frame number. As can be seen from the evaluation results, the present invention (fcsf 2) performs well in the target tracking task compared to the classical target tracking algorithm.
As shown in fig. 6, a comparison graph of the target tracking results in the actual video of the present invention and some excellent algorithms in recent years is shown, where (a) is a graph of the results of Human test video on OTB100 data set, (b) is a graph of the results of Walking test video on OTB100 data set, (c) is a graph of the results of Tiger test video on OTB50 data set, and (d) is a graph of the results of Dog test video on OTB50 data set. In general, compared with a classical target tracking algorithm, the fSCF2 tracking method has the best tracking effect, and has strong target representation capability due to the use of deep convolution characteristics, and the adaptive scale estimation mechanism and the adaptive model updating strategy with high confidence coefficient enable the method to accurately track the target under the adverse factor conditions of shielding, scale change, target deformation, rapid target movement and the like of the target.

Claims (6)

1. A scale self-adaptive target tracking method based on a depth feature kernel correlation filter is characterized by comprising the following steps:
step 1, inputting an initial position p of a target0Sum scale s0Setting the window size to be 2.0 times of the target initial bounding box;
step 2, according to the target position p of the t-1 th framet-1Obtaining a target area xt-1The size is the window size;
step 3, extracting a target area xt-1Is characterized by deep convolution ofPerforming fast Fourier transform to obtain characteristic spectrumWherein ^ represents a discrete Fourier transform;
step 4, according to the characteristic mapCompute kernel autocorrelation
Step 5, training a position and scale correlation filter;
step 6, according to the position p of the t-1 frame targett-1Obtaining a candidate region z of the target in the t-th frametThe size is the window size;
step 7, extracting candidate region ztAnd performing fast Fourier transform to obtain a feature mapWherein ^ represents a discrete Fourier transform;
step 8, according to the characteristic map of the previous frame of the targetComputing kernel cross-correlation
Step 9, respectively detecting the positions corresponding to the maximum values in the output maps of the position filter and the scale filter to determine the position p of the target in the current frametSum scale st；
And step 10, updating the kernel correlation filter by adopting a self-adaptive model updating strategy.
2. The method for tracking scale-adaptive targets based on the depth feature kernel correlation filter according to claim 1, wherein the method for extracting depth convolution features in steps 3 and 7 specifically comprises the following steps:
(3.1) preprocessing, scaling the window area I to the input size 224 × 224 specified by the convolutional neural network;
(3.2) extracting features, namely extracting feature maps of the 3 rd, 4 th and 5 th convolutional layers of the convolutional neural network;
and (3.3) carrying out bilinear interpolation, and upsampling the extracted 3-layer convolution characteristics to the same size.
3. The method for scale-adaptive target tracking based on depth feature kernel correlation filter according to claim 1, wherein the step 4 is to calculate kernel autocorrelationAnd the step 8 of calculating the kernel cross correlationThe specific method comprises the following steps:
(4.1) using a Gaussian kernel, the formula is as follows:
<mrow> <mi>k</mi> <mrow> <mo>(</mo> <mi>x</mi> <mo>,</mo> <msup> <mi>x</mi> <mo>&amp;prime;</mo> </msup> <mo>)</mo> </mrow> <mo>=</mo> <mi>exp</mi> <mrow> <mo>(</mo> <mo>-</mo> <mfrac> <mn>1</mn> <msup> <mi>&amp;sigma;</mi> <mn>2</mn> </msup> </mfrac> <mo>|</mo> <mo>|</mo> <mi>x</mi> <mo>-</mo> <msup> <mi>x</mi> <mo>&amp;prime;</mo> </msup> <mo>|</mo> <msup> <mo>|</mo> <mn>2</mn> </msup> <mo>)</mo> </mrow> </mrow>
wherein k (x, x ') represents a Gaussian kernel calculated by two feature maps x and x', exp (eta) represents an e index function, sigma is a standard deviation of the Gaussian function, the value of sigma is 0.5, | | |22-normal form representing a vector or matrix;
(4.2) calculating the kernel correlation, the formula is as follows:
wherein k isxx′Representing the nuclear correlation of the characteristic maps x and x', exp (eta) representing an e index function, sigma is the standard deviation of a Gaussian function, the value of sigma is 0.5, | | |. | tory2A 2-normal form representing a vector or matrix,represents the inverse of the discrete fourier transform, denotes the complex conjugate, denotes the discrete fourier transform, and ⊙ denotes the multiplication of the corresponding elements of the two matrices.
4. The method for tracking a scale-adaptive target based on a depth feature kernel correlation filter according to claim 1, wherein the training position and scale correlation filter of step 5 is specifically as follows:
respectively training 1 kernel correlation filter for each layer of feature spectrum according to the deep convolution features extracted in the step 3, and training a model by adopting the following formula:
<mrow> <msup> <mover> <mi>&amp;alpha;</mi> <mo>^</mo> </mover> <mrow> <mo>(</mo> <mi>l</mi> <mo>)</mo> </mrow> </msup> <mo>=</mo> <mfrac> <mover> <mi>y</mi> <mo>^</mo> </mover> <mrow> <msup> <mover> <mi>k</mi> <mo>^</mo> </mover> <mrow> <msup> <mi>xx</mi> <mrow> <mo>(</mo> <mi>l</mi> <mo>)</mo> </mrow> </msup> </mrow> </msup> <mo>+</mo> <mi>&amp;lambda;</mi> </mrow> </mfrac> </mrow>
wherein,representing a feature map convolved according to the l-th layer depthThe obtained correlation filter model is used for calculating the correlation filter model,representation feature mapRepresents discrete Fourier transform, lambda is a regularization parameter to prevent overfitting of the trained model, and lambda is 0.001.
5. The method according to claim 1, wherein the step 9 detects the position corresponding to the maximum value in the output maps of the position filter and the scale filter to determine the position p of the target in the current frametSum scale stThe method comprises the following steps:
(9.1) from the t-th frame image, with position pt-1And selecting candidate area z by window sizet,trans；
(9.2) extracting candidate region zt,trans3 layers depth convolution characteristic map
(9.3) calculating a position filter correlation output confidence coefficient map f for the first layer characteristic mapt,trans (l)：
Wherein f ist （l)The output corresponding map of the position filter representing the characteristic map of the l-th layer,representation feature mapAndthe cross-correlation of the kernels of (a),is the position filter trained and updated from the previous frame,represents the inverse of the discrete Fourier transform, represents the discrete Fourier transform, ⊙ represents the multiplication of the corresponding elements of the two matrices;
(9.4) outputting the corresponding map ft (3)Starting to estimate the target position from coarse to fine, the target position p of the l-th layert (l)For outputting the corresponding map ft,trans (l)The position corresponding to the maximum value;
(9.5) from the t-th frame image, with position ptSum scale st-1Extracting candidate region z of scale estimatet,sacleConstructing a scale pyramid;
(9.6) extracting gradient direction histogram feature of candidate regionComputing a scale filter correlation output confidence map ft,sacle；
(9.7) target dimension s detected at the t-th frametFor outputting the corresponding map ft,sacleThe maximum value corresponds to the scale.
6. The method for scale-adaptive target tracking based on the depth feature kernel correlation filter according to claim 1, wherein the step 10 updates the kernel correlation filter by using an adaptive model update strategy, specifically as follows:
(10.1) after completing the estimation of the target position and scale, calculating two tracking result confidence measure criteria, one of which is the peak value of the relevant corresponding output map:
fmax＝maxf
wherein f is a corresponding map output by the correlation between the kernel correlation filter and the candidate region, fmaxIs the peak of the map;
the other is the average peak correlation energy APCE:
<mrow> <mi>A</mi> <mi>P</mi> <mi>C</mi> <mi>E</mi> <mo>=</mo> <mfrac> <mrow> <mo>|</mo> <msub> <mi>f</mi> <mrow> <mi>m</mi> <mi>a</mi> <mi>x</mi> </mrow> </msub> <mo>-</mo> <msub> <mi>f</mi> <mrow> <mi>m</mi> <mi>i</mi> <mi>n</mi> </mrow> </msub> <msup> <mo>|</mo> <mn>2</mn> </msup> </mrow> <mrow> <mi>m</mi> <mi>e</mi> <mi>a</mi> <mi>n</mi> <mrow> <mo>(</mo> <msub> <mi>&amp;Sigma;</mi> <mrow> <mi>i</mi> <mo>,</mo> <mi>j</mi> </mrow> </msub> <msup> <mrow> <mo>(</mo> <msub> <mi>f</mi> <mrow> <mi>i</mi> <mo>,</mo> <mi>j</mi> </mrow> </msub> <mo>-</mo> <msub> <mi>f</mi> <mrow> <mi>m</mi> <mi>i</mi> <mi>n</mi> </mrow> </msub> <mo>)</mo> </mrow> <mn>2</mn> </msup> <mo>)</mo> </mrow> </mrow> </mfrac> </mrow>
wherein f ismaxAnd fminRespectively, of the output map fLarge and minimum values, mean () representing the averaging function, fi,jThe ith row and jth column of values representing the output map f;
(10.2) if fmaxAnd APCE are both greater than fmaxAnd APCE historical average value, updating the model, otherwise, not updating; for each depth convolution layer, updating using a linear interpolation method, the formula is as follows:
<mrow> <msup> <msub> <mover> <mi>&amp;alpha;</mi> <mo>^</mo> </mover> <mi>t</mi> </msub> <mrow> <mo>(</mo> <mi>l</mi> <mo>)</mo> </mrow> </msup> <mo>=</mo> <mrow> <mo>(</mo> <mn>1</mn> <mo>-</mo> <mi>&amp;eta;</mi> <mo>)</mo> </mrow> <msup> <msub> <mover> <mi>&amp;alpha;</mi> <mo>^</mo> </mover> <mrow> <mi>t</mi> <mo>-</mo> <mn>1</mn> </mrow> </msub> <mrow> <mo>(</mo> <mi>l</mi> <mo>)</mo> </mrow> </msup> <mo>+</mo> <mi>&amp;eta;</mi> <msup> <msub> <mover> <mi>&amp;alpha;</mi> <mo>^</mo> </mover> <mi>t</mi> </msub> <mrow> <mo>(</mo> <mi>l</mi> <mo>)</mo> </mrow> </msup> </mrow>2
<mrow> <msup> <msub> <mover> <mi>x</mi> <mo>^</mo> </mover> <mi>t</mi> </msub> <mrow> <mo>(</mo> <mi>l</mi> <mo>)</mo> </mrow> </msup> <mo>=</mo> <mrow> <mo>(</mo> <mn>1</mn> <mo>-</mo> <mi>&amp;eta;</mi> <mo>)</mo> </mrow> <msup> <msub> <mover> <mi>x</mi> <mo>^</mo> </mover> <mrow> <mi>t</mi> <mo>-</mo> <mn>1</mn> </mrow> </msub> <mrow> <mo>(</mo> <mi>l</mi> <mo>)</mo> </mrow> </msup> <mo>+</mo> <mi>&amp;eta;</mi> <msup> <msub> <mover> <mi>x</mi> <mo>^</mo> </mover> <mi>t</mi> </msub> <mrow> <mo>(</mo> <mi>l</mi> <mo>)</mo> </mrow> </msup> </mrow>
wherein,andthe feature map and the related filter of the previous frame of the ith layer are respectively represented, η is the learning rate, the larger η is, the faster model is updated, and η takes the value of 0.02.
CN201710355456.3A 2017-05-19 2017-05-19 Dimension self-adaption method for tracking target based on depth characteristic core correlation filter Pending CN107154024A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN201710355456.3A CN107154024A (en) 2017-05-19 2017-05-19 Dimension self-adaption method for tracking target based on depth characteristic core correlation filter

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201710355456.3A CN107154024A (en) 2017-05-19 2017-05-19 Dimension self-adaption method for tracking target based on depth characteristic core correlation filter

Publications (1)

Publication Number Publication Date
CN107154024A true CN107154024A (en) 2017-09-12

Family

ID=59794201

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201710355456.3A Pending CN107154024A (en) 2017-05-19 2017-05-19 Dimension self-adaption method for tracking target based on depth characteristic core correlation filter

Country Status (1)

Country Link
CN (1) CN107154024A (en)

Cited By (56)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN107644217A (en) * 2017-09-29 2018-01-30 中国科学技术大学 Method for tracking target based on convolutional neural networks and correlation filter
CN107730536A (en) * 2017-09-15 2018-02-23 北京飞搜科技有限公司 A kind of high speed correlation filtering object tracking method based on depth characteristic
CN108053424A (en) * 2017-12-15 2018-05-18 深圳云天励飞技术有限公司 Method for tracking target, device, electronic equipment and storage medium
CN108182388A (en) * 2017-12-14 2018-06-19 哈尔滨工业大学（威海） A kind of motion target tracking method based on image
CN108280808A (en) * 2017-12-15 2018-07-13 西安电子科技大学 The method for tracking target of correlation filter is exported based on structuring
CN108346159A (en) * 2018-01-28 2018-07-31 北京工业大学 A kind of visual target tracking method based on tracking-study-detection
CN108345885A (en) * 2018-01-18 2018-07-31 浙江大华技术股份有限公司 A kind of method and device of target occlusion detection
CN108550126A (en) * 2018-04-18 2018-09-18 长沙理工大学 A kind of adaptive correlation filter method for tracking target and system
CN108573499A (en) * 2018-03-16 2018-09-25 东华大学 A kind of visual target tracking method based on dimension self-adaption and occlusion detection
CN108665481A (en) * 2018-03-27 2018-10-16 西安电子科技大学 Multilayer depth characteristic fusion it is adaptive resist block infrared object tracking method
CN108734151A (en) * 2018-06-14 2018-11-02 厦门大学 Robust long-range method for tracking target based on correlation filtering and the twin network of depth
CN108830878A (en) * 2018-04-13 2018-11-16 上海大学 A kind of method for tracking target based on FPN neural network
CN108846345A (en) * 2018-06-06 2018-11-20 安徽大学 Moving object scale estimation method in monitoring scene
CN109035300A (en) * 2018-07-05 2018-12-18 桂林电子科技大学 A kind of method for tracking target based on depth characteristic Yu average peak correlation energy
CN109035290A (en) * 2018-07-16 2018-12-18 南京信息工程大学 A kind of track algorithm updating accretion learning based on high confidence level
CN109064491A (en) * 2018-04-12 2018-12-21 江苏省基础地理信息中心 A kind of nuclear phase pass filter tracking method of adaptive piecemeal
CN109146928A (en) * 2017-12-29 2019-01-04 西安电子科技大学 A kind of method for tracking target that Grads threshold judgment models update
CN109146917A (en) * 2017-12-29 2019-01-04 西安电子科技大学 A kind of method for tracking target of elasticity more new strategy
CN109166106A (en) * 2018-08-02 2019-01-08 山东大学 A kind of target detection aligning method and apparatus based on sliding window
CN109255304A (en) * 2018-08-17 2019-01-22 西安电子科技大学 Method for tracking target based on distribution field feature
CN109410246A (en) * 2018-09-25 2019-03-01 深圳市中科视讯智能系统技术有限公司 The method and device of vision tracking based on correlation filtering
CN109410251A (en) * 2018-11-19 2019-03-01 南京邮电大学 Method for tracking target based on dense connection convolutional network
CN109410247A (en) * 2018-10-16 2019-03-01 中国石油大学(华东) A kind of video tracking algorithm of multi-template and adaptive features select
CN109461172A (en) * 2018-10-25 2019-03-12 南京理工大学 Manually with the united correlation filtering video adaptive tracking method of depth characteristic
CN109584271A (en) * 2018-11-15 2019-04-05 西北工业大学 High speed correlation filtering tracking based on high confidence level more new strategy
CN109741366A (en) * 2018-11-27 2019-05-10 昆明理工大学 A kind of correlation filtering method for tracking target merging multilayer convolution feature
CN109785360A (en) * 2018-12-18 2019-05-21 南京理工大学 A kind of adaptive method for real time tracking excavated based on online sample
CN109801311A (en) * 2019-01-31 2019-05-24 长安大学 A kind of visual target tracking method based on depth residual error network characterization
CN109858455A (en) * 2019-02-18 2019-06-07 南京航空航天大学 A kind of piecemeal detection scale adaptive tracking method for circular target
CN109859244A (en) * 2019-01-22 2019-06-07 西安微电子技术研究所 A kind of visual tracking method based on convolution sparseness filtering
CN109858454A (en) * 2019-02-15 2019-06-07 东北大学 One kind being based on dual model self-adaptive kernel correlation filtering method for tracing
CN109858326A (en) * 2018-12-11 2019-06-07 中国科学院自动化研究所 Based on classification semantic Weakly supervised online visual tracking method and system
CN109886996A (en) * 2019-01-15 2019-06-14 东华大学 A kind of visual pursuit optimization method
CN109934098A (en) * 2019-01-24 2019-06-25 西北工业大学 A kind of video camera intelligence system and its implementation with secret protection
CN110033006A (en) * 2019-04-04 2019-07-19 中设设计集团股份有限公司 Vehicle detecting and tracking method based on color characteristic Nonlinear Dimension Reduction
CN110197126A (en) * 2019-05-06 2019-09-03 深圳岚锋创视网络科技有限公司 A kind of target tracking method, device and portable terminal
CN110211149A (en) * 2018-12-25 2019-09-06 湖州云通科技有限公司 A kind of dimension self-adaption nuclear phase pass filter tracking method based on context-aware
CN110414439A (en) * 2019-07-30 2019-11-05 武汉理工大学 Anti- based on multi-peak detection blocks pedestrian tracting method
CN110544267A (en) * 2019-07-24 2019-12-06 中国地质大学（武汉） correlation filtering tracking method for self-adaptive selection characteristics
CN110555870A (en) * 2019-09-09 2019-12-10 北京理工大学 DCF tracking confidence evaluation and classifier updating method based on neural network
CN110633595A (en) * 2018-06-21 2019-12-31 北京京东尚科信息技术有限公司 Target detection method and device by utilizing bilinear interpolation
CN110689559A (en) * 2019-09-30 2020-01-14 长安大学 Visual target tracking method based on dense convolutional network characteristics
CN110807473A (en) * 2019-10-12 2020-02-18 浙江大华技术股份有限公司 Target detection method, device and computer storage medium
CN110889863A (en) * 2019-09-03 2020-03-17 河南理工大学 Target tracking method based on target perception correlation filtering
CN111161323A (en) * 2019-12-31 2020-05-15 北京理工大学重庆创新中心 Complex scene target tracking method and system based on correlation filtering
CN111192288A (en) * 2018-11-14 2020-05-22 天津大学青岛海洋技术研究院 Target tracking algorithm based on deformation sample generation network
CN111221770A (en) * 2019-12-31 2020-06-02 中国船舶重工集团公司第七一七研究所 Kernel correlation filtering target tracking method and system
CN111428740A (en) * 2020-02-28 2020-07-17 深圳壹账通智能科技有限公司 Detection method and device for network-shot photo, computer equipment and storage medium
CN111476819A (en) * 2020-03-19 2020-07-31 重庆邮电大学 Long-term target tracking method based on multi-correlation filtering model
CN111507999A (en) * 2019-01-30 2020-08-07 北京四维图新科技股份有限公司 FDSST algorithm-based target tracking method and device
CN111696132A (en) * 2020-05-15 2020-09-22 深圳市优必选科技股份有限公司 Target tracking method and device, computer readable storage medium and robot
CN112053386A (en) * 2020-08-31 2020-12-08 西安电子科技大学 Target tracking method based on depth convolution characteristic self-adaptive integration
CN112200833A (en) * 2020-09-17 2021-01-08 天津城建大学 Relevant filtering video tracking algorithm based on residual error network and short-term visual memory
CN112560695A (en) * 2020-12-17 2021-03-26 中国海洋大学 Underwater target tracking method, system, storage medium, equipment, terminal and application
CN113379804A (en) * 2021-07-12 2021-09-10 闽南师范大学 Unmanned aerial vehicle target tracking method, terminal equipment and storage medium
CN114708300A (en) * 2022-03-02 2022-07-05 北京理工大学 Anti-blocking self-adaptive target tracking method and system

Citations (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
WO2015163830A1 (en) * 2014-04-22 2015-10-29 Aselsan Elektronik Sanayi Ve Ticaret Anonim Sirketi Target localization and size estimation via multiple model learning in visual tracking
CN106570486A (en) * 2016-11-09 2017-04-19 华南理工大学 Kernel correlation filtering target tracking method based on feature fusion and Bayesian classification

Patent Citations (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
WO2015163830A1 (en) * 2014-04-22 2015-10-29 Aselsan Elektronik Sanayi Ve Ticaret Anonim Sirketi Target localization and size estimation via multiple model learning in visual tracking
CN106570486A (en) * 2016-11-09 2017-04-19 华南理工大学 Kernel correlation filtering target tracking method based on feature fusion and Bayesian classification

Non-Patent Citations (1)

* Cited by examiner, † Cited by third party
Title
张雷: "复杂场景下实时目标跟踪算法及实现技术研究", 《中国博士学位论文全文数据库 信息科技辑》 *

Cited By (92)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN107730536A (en) * 2017-09-15 2018-02-23 北京飞搜科技有限公司 A kind of high speed correlation filtering object tracking method based on depth characteristic
CN107730536B (en) * 2017-09-15 2020-05-12 苏州飞搜科技有限公司 High-speed correlation filtering object tracking method based on depth features
CN107644217B (en) * 2017-09-29 2020-06-26 中国科学技术大学 Target tracking method based on convolutional neural network and related filter
CN107644217A (en) * 2017-09-29 2018-01-30 中国科学技术大学 Method for tracking target based on convolutional neural networks and correlation filter
CN108182388A (en) * 2017-12-14 2018-06-19 哈尔滨工业大学（威海） A kind of motion target tracking method based on image
CN108053424A (en) * 2017-12-15 2018-05-18 深圳云天励飞技术有限公司 Method for tracking target, device, electronic equipment and storage medium
CN108280808A (en) * 2017-12-15 2018-07-13 西安电子科技大学 The method for tracking target of correlation filter is exported based on structuring
CN108053424B (en) * 2017-12-15 2020-06-16 深圳云天励飞技术有限公司 Target tracking method and device, electronic equipment and storage medium
CN109146928A (en) * 2017-12-29 2019-01-04 西安电子科技大学 A kind of method for tracking target that Grads threshold judgment models update
CN109146917A (en) * 2017-12-29 2019-01-04 西安电子科技大学 A kind of method for tracking target of elasticity more new strategy
CN109146928B (en) * 2017-12-29 2021-09-24 西安电子科技大学 Target tracking method for updating gradient threshold judgment model
CN109146917B (en) * 2017-12-29 2020-07-28 西安电子科技大学 Target tracking method for elastic updating strategy
CN108345885A (en) * 2018-01-18 2018-07-31 浙江大华技术股份有限公司 A kind of method and device of target occlusion detection
CN108346159A (en) * 2018-01-28 2018-07-31 北京工业大学 A kind of visual target tracking method based on tracking-study-detection
CN108346159B (en) * 2018-01-28 2021-10-15 北京工业大学 Tracking-learning-detection-based visual target tracking method
CN108573499A (en) * 2018-03-16 2018-09-25 东华大学 A kind of visual target tracking method based on dimension self-adaption and occlusion detection
CN108573499B (en) * 2018-03-16 2021-04-02 东华大学 Visual target tracking method based on scale self-adaption and occlusion detection
CN108665481A (en) * 2018-03-27 2018-10-16 西安电子科技大学 Multilayer depth characteristic fusion it is adaptive resist block infrared object tracking method
CN108665481B (en) * 2018-03-27 2022-05-31 西安电子科技大学 Self-adaptive anti-blocking infrared target tracking method based on multi-layer depth feature fusion
CN109064491A (en) * 2018-04-12 2018-12-21 江苏省基础地理信息中心 A kind of nuclear phase pass filter tracking method of adaptive piecemeal
CN108830878A (en) * 2018-04-13 2018-11-16 上海大学 A kind of method for tracking target based on FPN neural network
CN108830878B (en) * 2018-04-13 2021-02-23 上海大学 Target tracking method based on FPN neural network
CN108550126A (en) * 2018-04-18 2018-09-18 长沙理工大学 A kind of adaptive correlation filter method for tracking target and system
CN108846345B (en) * 2018-06-06 2021-09-17 安徽大学 Moving object scale estimation method in monitoring scene
CN108846345A (en) * 2018-06-06 2018-11-20 安徽大学 Moving object scale estimation method in monitoring scene
CN108734151B (en) * 2018-06-14 2020-04-14 厦门大学 Robust long-range target tracking method based on correlation filtering and depth twin network
CN108734151A (en) * 2018-06-14 2018-11-02 厦门大学 Robust long-range method for tracking target based on correlation filtering and the twin network of depth
CN110633595B (en) * 2018-06-21 2022-12-02 北京京东尚科信息技术有限公司 Target detection method and device by utilizing bilinear interpolation
CN110633595A (en) * 2018-06-21 2019-12-31 北京京东尚科信息技术有限公司 Target detection method and device by utilizing bilinear interpolation
CN109035300B (en) * 2018-07-05 2021-03-26 桂林电子科技大学 Target tracking method based on depth feature and average peak correlation energy
CN109035300A (en) * 2018-07-05 2018-12-18 桂林电子科技大学 A kind of method for tracking target based on depth characteristic Yu average peak correlation energy
CN109035290A (en) * 2018-07-16 2018-12-18 南京信息工程大学 A kind of track algorithm updating accretion learning based on high confidence level
CN109166106A (en) * 2018-08-02 2019-01-08 山东大学 A kind of target detection aligning method and apparatus based on sliding window
CN109255304A (en) * 2018-08-17 2019-01-22 西安电子科技大学 Method for tracking target based on distribution field feature
CN109255304B (en) * 2018-08-17 2021-07-27 西安电子科技大学 Target tracking method based on distribution field characteristics
CN109410246B (en) * 2018-09-25 2021-06-11 杭州视语智能视觉系统技术有限公司 Visual tracking method and device based on correlation filtering
CN109410246A (en) * 2018-09-25 2019-03-01 深圳市中科视讯智能系统技术有限公司 The method and device of vision tracking based on correlation filtering
CN109410247A (en) * 2018-10-16 2019-03-01 中国石油大学(华东) A kind of video tracking algorithm of multi-template and adaptive features select
CN109461172A (en) * 2018-10-25 2019-03-12 南京理工大学 Manually with the united correlation filtering video adaptive tracking method of depth characteristic
CN111192288B (en) * 2018-11-14 2023-08-04 天津大学青岛海洋技术研究院 Target tracking algorithm based on deformation sample generation network
CN111192288A (en) * 2018-11-14 2020-05-22 天津大学青岛海洋技术研究院 Target tracking algorithm based on deformation sample generation network
CN109584271A (en) * 2018-11-15 2019-04-05 西北工业大学 High speed correlation filtering tracking based on high confidence level more new strategy
CN109410251A (en) * 2018-11-19 2019-03-01 南京邮电大学 Method for tracking target based on dense connection convolutional network
CN109741366A (en) * 2018-11-27 2019-05-10 昆明理工大学 A kind of correlation filtering method for tracking target merging multilayer convolution feature
CN109858326A (en) * 2018-12-11 2019-06-07 中国科学院自动化研究所 Based on classification semantic Weakly supervised online visual tracking method and system
CN109785360A (en) * 2018-12-18 2019-05-21 南京理工大学 A kind of adaptive method for real time tracking excavated based on online sample
CN110211149A (en) * 2018-12-25 2019-09-06 湖州云通科技有限公司 A kind of dimension self-adaption nuclear phase pass filter tracking method based on context-aware
CN110211149B (en) * 2018-12-25 2022-08-12 湖州云通科技有限公司 Scale self-adaptive kernel correlation filtering tracking method based on background perception
CN109886996B (en) * 2019-01-15 2023-06-06 东华大学 Visual tracking optimization method
CN109886996A (en) * 2019-01-15 2019-06-14 东华大学 A kind of visual pursuit optimization method
CN109859244A (en) * 2019-01-22 2019-06-07 西安微电子技术研究所 A kind of visual tracking method based on convolution sparseness filtering
CN109859244B (en) * 2019-01-22 2022-07-08 西安微电子技术研究所 Visual tracking method based on convolution sparse filtering
CN109934098A (en) * 2019-01-24 2019-06-25 西北工业大学 A kind of video camera intelligence system and its implementation with secret protection
CN111507999B (en) * 2019-01-30 2023-07-18 北京四维图新科技股份有限公司 Target tracking method and device based on FDSST algorithm
CN111507999A (en) * 2019-01-30 2020-08-07 北京四维图新科技股份有限公司 FDSST algorithm-based target tracking method and device
CN109801311B (en) * 2019-01-31 2021-07-16 长安大学 Visual target tracking method based on depth residual error network characteristics
CN109801311A (en) * 2019-01-31 2019-05-24 长安大学 A kind of visual target tracking method based on depth residual error network characterization
CN109858454A (en) * 2019-02-15 2019-06-07 东北大学 One kind being based on dual model self-adaptive kernel correlation filtering method for tracing
CN109858454B (en) * 2019-02-15 2023-04-07 东北大学 Adaptive kernel correlation filtering tracking method based on dual models
CN109858455A (en) * 2019-02-18 2019-06-07 南京航空航天大学 A kind of piecemeal detection scale adaptive tracking method for circular target
CN110033006A (en) * 2019-04-04 2019-07-19 中设设计集团股份有限公司 Vehicle detecting and tracking method based on color characteristic Nonlinear Dimension Reduction
CN110197126A (en) * 2019-05-06 2019-09-03 深圳岚锋创视网络科技有限公司 A kind of target tracking method, device and portable terminal
CN110544267B (en) * 2019-07-24 2022-03-15 中国地质大学（武汉） Correlation filtering tracking method for self-adaptive selection characteristics
CN110544267A (en) * 2019-07-24 2019-12-06 中国地质大学（武汉） correlation filtering tracking method for self-adaptive selection characteristics
CN110414439A (en) * 2019-07-30 2019-11-05 武汉理工大学 Anti- based on multi-peak detection blocks pedestrian tracting method
CN110414439B (en) * 2019-07-30 2022-03-15 武汉理工大学 Anti-blocking pedestrian tracking method based on multi-peak detection
CN110889863A (en) * 2019-09-03 2020-03-17 河南理工大学 Target tracking method based on target perception correlation filtering
CN110889863B (en) * 2019-09-03 2023-03-24 河南理工大学 Target tracking method based on target perception correlation filtering
CN110555870B (en) * 2019-09-09 2021-07-27 北京理工大学 DCF tracking confidence evaluation and classifier updating method based on neural network
CN110555870A (en) * 2019-09-09 2019-12-10 北京理工大学 DCF tracking confidence evaluation and classifier updating method based on neural network
CN110689559A (en) * 2019-09-30 2020-01-14 长安大学 Visual target tracking method based on dense convolutional network characteristics
CN110689559B (en) * 2019-09-30 2022-08-12 长安大学 Visual target tracking method based on dense convolutional network characteristics
CN110807473A (en) * 2019-10-12 2020-02-18 浙江大华技术股份有限公司 Target detection method, device and computer storage medium
CN110807473B (en) * 2019-10-12 2023-01-03 浙江大华技术股份有限公司 Target detection method, device and computer storage medium
CN111161323B (en) * 2019-12-31 2023-11-28 北京理工大学重庆创新中心 Complex scene target tracking method and system based on correlation filtering
CN111221770A (en) * 2019-12-31 2020-06-02 中国船舶重工集团公司第七一七研究所 Kernel correlation filtering target tracking method and system
CN111161323A (en) * 2019-12-31 2020-05-15 北京理工大学重庆创新中心 Complex scene target tracking method and system based on correlation filtering
WO2021169625A1 (en) * 2020-02-28 2021-09-02 深圳壹账通智能科技有限公司 Method and apparatus for detecting reproduced network photograph, computer device, and storage medium
CN111428740A (en) * 2020-02-28 2020-07-17 深圳壹账通智能科技有限公司 Detection method and device for network-shot photo, computer equipment and storage medium
CN111476819A (en) * 2020-03-19 2020-07-31 重庆邮电大学 Long-term target tracking method based on multi-correlation filtering model
WO2021227519A1 (en) * 2020-05-15 2021-11-18 深圳市优必选科技股份有限公司 Target tracking method and apparatus, and computer-readable storage medium and robot
CN111696132B (en) * 2020-05-15 2023-12-29 深圳市优必选科技股份有限公司 Target tracking method, device, computer readable storage medium and robot
CN111696132A (en) * 2020-05-15 2020-09-22 深圳市优必选科技股份有限公司 Target tracking method and device, computer readable storage medium and robot
CN112053386B (en) * 2020-08-31 2023-04-18 西安电子科技大学 Target tracking method based on depth convolution characteristic self-adaptive integration
CN112053386A (en) * 2020-08-31 2020-12-08 西安电子科技大学 Target tracking method based on depth convolution characteristic self-adaptive integration
CN112200833A (en) * 2020-09-17 2021-01-08 天津城建大学 Relevant filtering video tracking algorithm based on residual error network and short-term visual memory
CN112560695B (en) * 2020-12-17 2023-03-24 中国海洋大学 Underwater target tracking method, system, storage medium, equipment, terminal and application
CN112560695A (en) * 2020-12-17 2021-03-26 中国海洋大学 Underwater target tracking method, system, storage medium, equipment, terminal and application
CN113379804A (en) * 2021-07-12 2021-09-10 闽南师范大学 Unmanned aerial vehicle target tracking method, terminal equipment and storage medium
CN113379804B (en) * 2021-07-12 2023-05-09 闽南师范大学 Unmanned aerial vehicle target tracking method, terminal equipment and storage medium
CN114708300A (en) * 2022-03-02 2022-07-05 北京理工大学 Anti-blocking self-adaptive target tracking method and system
CN114708300B (en) * 2022-03-02 2024-07-23 北京理工大学 Anti-shielding self-adaptive target tracking method and system

Similar Documents

Publication Publication Date Title
CN107154024A (en) Dimension self-adaption method for tracking target based on depth characteristic core correlation filter
CN108665481B (en) Self-adaptive anti-blocking infrared target tracking method based on multi-layer depth feature fusion
CN111354017B (en) Target tracking method based on twin neural network and parallel attention module
WO2022036777A1 (en) Method and device for intelligent estimation of human body movement posture based on convolutional neural network
CN107316316A (en) The method for tracking target that filtering technique is closed with nuclear phase is adaptively merged based on multiple features
CN112184752A (en) Video target tracking method based on pyramid convolution
CN109741366B (en) Related filtering target tracking method fusing multilayer convolution characteristics
CN111462191B (en) Non-local filter unsupervised optical flow estimation method based on deep learning
CN110175649B (en) Rapid multi-scale estimation target tracking method for re-detection
CN107369166A (en) A kind of method for tracking target and system based on multiresolution neutral net
CN112364931B (en) Few-sample target detection method and network system based on meta-feature and weight adjustment
CN108053419A (en) Inhibited and the jamproof multiscale target tracking of prospect based on background
CN111080675A (en) Target tracking method based on space-time constraint correlation filtering
CN109461172A (en) Manually with the united correlation filtering video adaptive tracking method of depth characteristic
CN108427921A (en) A kind of face identification method based on convolutional neural networks
CN104574445A (en) Target tracking method and device
CN110120065B (en) Target tracking method and system based on hierarchical convolution characteristics and scale self-adaptive kernel correlation filtering
CN114565655B (en) Depth estimation method and device based on pyramid segmentation attention
CN106338733A (en) Forward-looking sonar object tracking method based on frog-eye visual characteristic
CN107452022A (en) A kind of video target tracking method
CN112949493A (en) Lane line detection method and system combining semantic segmentation and attention mechanism
CN116524062B (en) Diffusion model-based 2D human body posture estimation method
CN110245587B (en) Optical remote sensing image target detection method based on Bayesian transfer learning
CN107944354A (en) A kind of vehicle checking method based on deep learning
CN110660080A (en) Multi-scale target tracking method based on learning rate adjustment and fusion of multilayer convolution features

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
RJ01 Rejection of invention patent application after publication
RJ01 Rejection of invention patent application after publication

Application publication date: 20170912