CN104463192A - Dark environment video target real-time tracking method based on textural features - Google Patents

Abstract

The invention discloses a dark environment video target real-time tracking method based on textural features. A multi-scale rectangular filter serves as a signal sampling matrix, a sparse and random gauss matrix serves as a compression sensing matrix, and therefore sample features can be quickly extracted through a vector integral diagram algorithm; by the adoption of template trimming method, redundancy computation can be effectively reduced in the vector integral diagram step. The method utilizes an ULBP operator in a rotation invariant mode to extract the features, is suitable for target tracking of targets which possibly rotate and deform at dark night and under a poor underground lighting condition, is high in recognition rate and provides reliable results for target tracking.

Description

Based on the dark situation video object method for real time tracking of textural characteristics

Technical field

The present invention relates to a kind of dark situation video object method for real time tracking based on textural characteristics, belong to image pattern recognition field.

Background technology

Computer vision target tracking domain generally adopts detection formula to follow the tracks of framework, and tracing task, by generating a small amount of positive negative sample on-line training sorter, is converted into Detection task by this framework.This is because object detection field achieves major progress, classifier technique is also constantly made progress by large quantity research, effectively ensure that the success ratio of tracking.Detection task needs to carry out feature extraction to the sample collected, with the feature of reflected sample, just sample classification and differentiation can be carried out, traditional feature extracting method needs dependence experience to construct, K.H.Zhang etc. propose a kind of feature extracting method based on compressed sensing (CompressiveTracking), by by broad sense Haar feature and a series of multi-scale filtering device convolution to ensure the multiple dimensioned character of feature, recycling Random sparseness Gaussian matrix carries out dimensionality reduction to ensure the live effect of tracking to feature.But there is the characteristic to illumination brightness, target rotational sensitive in broad sense Haar feature, the present invention uses invariable rotary pattern ULBP operator to improve the flow process of feature extraction, on the basis ensureing real-time and stability, make target tracking algorism can adapt to all kinds of scenes easily causing track rejection such as low-light (level), target rotation, illumination variation.

Summary of the invention

The problem of target following in extreme illumination scene can not be processed in order to overcome existing track algorithm.The present invention proposes a kind of object real-time tracking method based on textural characteristics of the particular surroundingss such as applicable down-hole, night, the method utilizes invariable rotary pattern ULBP operator to carry out texture feature extraction, the feature after extraction is made to contain the texture information of abundant sample, utilize texture information to the characteristic of illumination-insensitive, enable tracker in dim environment, reach higher tracking success ratio.

The invention discloses a kind of dark situation video object method for real time tracking based on textural characteristics, comprise initial phase and target tracking stage, described initial phase comprises the following steps:

1) the compute sparse sampling matrix Θ when initialization:

A) signal sampling matrix Φ is calculated;

B) compute sparse perception matrix Ψ;

C) compute sparse sampling matrix Θ, wherein Θ=Ψ Φ;

2) establishment one is by two classification Naive Bayes Classifier H (x) of 50 Bayes classifier cascades, each Weak Classifier h _c(x _c) be all based on representing two normal distributions of positive label y=0 and negative label y=1 wherein (μ _{y, c}, σ _{y, c}) represent that label is the parameter value of the Normal Discrimination curve of the Weak Classifier that the c dimensional feature of y is corresponding;

Described target tracking stage comprises the following steps:

1) to the target detection of the kth frame of video

A) with the target O that kth-1 frame traces into _k-1centered by carry out candidate samples collection, in kth frame, collect n _yindividual Euclidean distance meets $z_y=\left\{z\right|0\≤{\left|\right|z-O_{k-1}\left|\right|}^{l2}{\≤r}_y^{+}\}$ Candidate samples;

B) minimum rectangular area ∪ z (the z ∈ z comprising whole candidate samples is calculated _y), gray processing is carried out successively to this rectangular region image sheet, invariable rotary pattern ULBP encodes, vectorial integration, finally obtains vector product component I;

C) with the diagonal line of the nonzero element in sparse sampling matrix Θ for scale, from vector product component I, extract compressed encoding eigenwert z → x (z ∈ z of each candidate samples with diagonal line subtraction _y);

D) by the compressed encoding eigenwert x of each candidate samples _rthe classifier calculated classification score that input kth-1 frame trains the x that classification score is maximum _reven if the target O that r corresponding sample kth frame traces into _k;

2) sorter of kth frame upgrades

A) with the target O that kth frame traces into _kcentered by carry out positive and negative sample collection, in kth frame, collect n ₁individual Euclidean distance meets positive sample, in kth frame, collect n ₀individual Euclidean distance meets $z_0=\left\{z\right|r_0^{-}\≤{\left|\right|z-O_k\left|\right|}^{l2}{\≤r}_0^{+}\}$ Negative sample;

B) minimum rectangular area ∪ z (the z ∈ z comprising all positive negative samples is calculated ₁∪ z ₀), gray processing is carried out successively to this rectangular region image sheet, invariable rotary pattern ULBP encodes, vectorial integration, finally obtains vector product component I;

C) with the diagonal line of the nonzero element in sparse sampling matrix Θ for scale, from vector product component I, extract compressed encoding eigenwert z → x (z ∈ z of each positive negative sample with diagonal line subtraction ₁∪ z ₀);

D) sorter is upgraded

$\left\{\begin{array}{c}{\mathrm{\μ}}_{y,c}^{\′}\←(1-\mathrm{\λ}){\mathrm{\μ}}_{y,c}+\mathrm{\λ}{\mathrm{EX}}_{y,c}\\ {\mathrm{\σ}}_{y,c}^{\′}\←{[(1-\mathrm{\λ}){\mathrm{\σ}}_{y,c}^2+\mathrm{\λ}{\mathrm{DX}}_{y,c}+\mathrm{\λ}(1-\mathrm{\λ}){({\mathrm{\μ}}_{y,c}-{\mathrm{EX}}_{y,c})}^2]}^{1/2}\end{array}\right.$

Wherein with the average for positive sample (y=1) and negative sample (y=0) and variance respectively.

The present invention further discloses described target tracking stage by gray-scale map I _grayto uniform pattern ULBP code pattern I _rULBPcoding method comprises the following steps:

1) pending pixel is as central pixel point, its gray-scale value g ₀with p the pixel g that distance is R _i∈ Γ _pthe difference of gray-scale value carry out binaryzation, and connect into the binary number of a p position, if the transition times U of 0 ~ 1 or 1 ~ 0 is not more than 2 in this binary number, then with the number of element 1, the uniform pattern ULBP as this pending pixel encodes, wherein $U=\underset{i=1}{\overset{p}{\mathrm{\Σ}}}1-\mathrm{\δ}[\mathrm{\ϵ}(g_{\mathrm{mod}(i+1,p)}-g_0)-\mathrm{\ϵ}(g_i-g_0)];$

The present invention further discloses described target tracking stage by uniform pattern ULBP code pattern I _rULBPintegration method to vector product component I comprises the following steps:

1) statistics with histogram, each coding occurrence number of 9 dimension statistics with histogram of structure corresponding 0 ~ 8;

2) longitudinally cumulative, its step is

A) carry out flattening by row to H, after flattening, obtain a dimensional vector V _c;

B) to V _cadd up, the cumulative dimensional vector obtained meets

C) to V _{Σ C}carry out fractureing by row, obtain the equal-sized image with H, count H _i;

3) laterally cumulative, its step is

A) to H _iflatten by row, after flattening, obtain one dimension row vector V _r;

B) to V _radd up, the cumulative dimensional vector obtained meets

C) to V _{Σ R}carry out fractureing by row, obtain the equal-sized image with H, count H _ii;

H _iibe the image I after process, in H, the statistic histogram of anyon matrix all can by H _iicarry out diagonal line subtraction to try to achieve.

Accompanying drawing explanation

Below in conjunction with the drawings and specific embodiments, the present invention is described in further detail.

Fig. 1 is the dark situation video object real-time follow-up process flow diagram based on textural characteristics;

Fig. 2 is sparse sampling matrix Θ and sample convolution schematic diagram;

Fig. 3 is the ROI region figure after coding;

Embodiment

Below in conjunction with Figure of description, the specific embodiment of the invention is described in detail, first the basic procedure of the dark situation video object method for real time tracking based on textural characteristics is described.With reference to Fig. 1, process is divided into initial phase, target tracking stage, the sorter more new stage, and its concrete steps are as follows, initial phase:

1) compute sparse sampling matrix Θ;

A) the long-pending Monte Carlo simulation utilizing signal sampling matrix Φ to be multiplied with sparse perception matrix Ψ calculates, with initial target O ₁rectangle is action scope, generates the rectangle frame of 2 ~ 3 random sizes of random site, and these rectangle frames are contained in O in wanting ₁, using a line nonzero element of these rectangle frames as Θ;

B) step 1a is repeated) d time, with reference to Fig. 2, obtain whole nonzero elements that the d of Θ is capable;

2) generate d and tie up Naive Bayes Classifier h (x);

A) generate one two classification Bayes classifier as a Weak Classifier, its positive label Critical curve parameter is μ _{1, c}=0, σ _{1, c}=1, its negative label Critical curve parameter is μ _{0, c}=0, σ _{0, c}=1;

B) step 2a is repeated) d time, obtain d the cascade Weak Classifier h of h (x) _c(x _c), wherein c=1,2 ..., d;

3) the calculating masterplate that gray-scale value is encoded to invariable rotary pattern ULBP is created;

Radius is adopted to be that the invariable rotary pattern ULBP operator of 1 is as the calculating masterplate of RULBP, generating the length comprising 0 ~ 255 is the array of 256, each element subscript value in array is converted into binary digit, the transition times of adding up this scale-of-two subscript value adjacent bit position 0 ~ 1 or 1 ~ 0 counts U, if U≤2, the number of times then bit 1 in this scale-of-two subscript value occurred as element value, otherwise using 0 as element value, the length after operation be 256 array comprise 0 ~ 89 kinds of codings;

Target tracking stage:

1) to the target detection of the kth frame of video

A) with the target O that kth-1 frame traces into _k-1top left corner apex centered by, calculate its distance meet all pixel { p _y, with { p _ybe top left corner apex, with O _k-1size sized by, the rectangle of gained is candidate samples z _y, count $z_y=\left\{z\right|0\≤{\left|\right|z-O_{k-1}\left|\right|}^{l2}{\≤r}_y^{+}\};$

B) comprise the minimum rectangular area ROI of whole candidate samples, with reference to Fig. 3, its computing method are ∪ z (z ∈ z _y), rectangle union operator ∪ is O (l ₁, r ₁, t ₁, b ₁) ∪ O (l ₂, r ₂, t ₂, b ₂)=O (max (l ₁, l ₂), min (r ₁, r ₂), max (t ₁, t ₂), min (b ₁, b ₂));

C) to the image sheet feature coding that ROI comprises, for pixel p each in ROI image sheet, its adjacent 8 pixels, with right pixel point for starting point, according to counterclockwise arrangement, 8 points after sequence are expressed as q successively ₁~ q ₈, have compare q successively _lwith the gray-scale value size of p, by 8 times relatively after logical value connect be 8 bits, the invariable rotary pattern ULBP as this pixel encodes;

D) ROI after coding counts matrix H and carries out vectorial integration, carries out flattening by row, obtain a dimensional vector V after flattening to H _c, to V _cadd up, the cumulative dimensional vector obtained meets to V _{Σ C}carry out fractureing by row, obtain the equal-sized image with H, count H _i, to H _iflatten by row, after flattening, obtain one dimension row vector V _r, to V _radd up, the cumulative dimensional vector obtained meets to V _{Σ R}carry out fractureing by row, obtain the equal-sized image with H, count H _ii;

2 ~ 3 nonzero elements of the often row of the sparse sampling matrix Θ e) generated by initialization procedure, wherein each element is all rectangular filters, and these wave filters are to candidate samples z _rcarry out results added that filtering the obtains one dimension x as feature _{r, c}, all carry out same operation by capable for the d of Θ, namely obtain candidate samples z _rd dimensional feature x _r=(x _{r, 1}, x _{r, 2}..., x _{r, d});

F) whole candidate samples z is calculated _r∈ z _yfeature utilize the Naive Bayes Classifier h (x that kth-1 frame trains _r; K-1) feature of each candidate samples classified and calculate classification score $h(x_r;k-1)=\underset{c=1}{\overset{d}{\mathrm{\Σ}}}h(x_{r,c};k-1)=\underset{c=1}{\overset{d}{\mathrm{\Σ}}}\log \frac{p\left(x_{r,c}\right|y=1)}{p\left(x_{r,c}\right|y=0)},$ Wherein $p\left(x_{r,c}\right|y)=\frac{1}{\sqrt{2\mathrm{\π}}{\mathrm{\σ}}_{y,c}}{e}^{\frac{(x_{r,c}-{\mathrm{\μ}}_{y,c})}{2{\mathrm{\σ}}_{y,c}^2}},$ Using the target O that candidate samples maximum for classification score traces into as kth frame _k;

2) sorter of kth frame upgrades

A) with the target O that kth frame traces into _ktop left corner apex centered by, calculate its distance meet all pixel { p ₁, with { p ₁be upper left corner fixed point, with O _ksize sized by, the rectangle of gained is positive sample z ₁, count $z_1=\left\{z\right|0\≤{\left|\right|z-O_{k-1}\left|\right|}^{l2}{\≤r}_y^{+}\};$

B) with the target O that kth frame traces into _ktop left corner apex centered by, calculate its distance meet all pixel { p ₀, with { p ₀be upper left corner fixed point, with O _ksize sized by, the rectangle of gained is negative sample z ₀, count $z_0=\left\{z\right|r_0^{-}\≤{\left|\right|z-O_{k-1}\left|\right|}^{l2}{\≤r}_0^{+}\};$

C) comprise the minimum rectangular area ROI of all positive negative samples, with reference to Fig. 2, its computing method are ∪ z (z ∈ z ₁∪ z ₀), the step 1b of rectangle union operator ∪ and the target detection of target tracking stage kth frame) identical;

D) the step 1c of the target detection of target tracking stage kth frame is repeated);

E) the step 1d of the target detection of target tracking stage kth frame is repeated);

2 ~ 3 nonzero elements of the often row of the sparse sampling matrix Θ f) generated by initialization procedure, wherein each element is all rectangular filters, and these wave filters align negative sample z _rcarry out results added that filtering the obtains one dimension x as feature _{r, c}, all carry out same operation by capable for the d of Θ, namely obtain positive negative sample z _rd dimensional feature x _r=(x _{r, 1}, x _{r, 2}..., x _{r, d});

G) all positive sample z are calculated _r∈ z ₁feature wherein each positive sample z _rfeature be all d tie up x _r=(x _{r, 1}, x _{r, 2}..., x _{r, d}), solve the average of all positive sample characteristics and variance ${\mathrm{DX}}_{1,c}=\frac{1}{n_1}\underset{r=1}{\overset{n_1}{\mathrm{\Σ}}}{x}_{1,r,c}^2-{\left({\mathrm{EX}}_{1,c}\right)}^2,$ Wherein c=1,2 ..., d;

H) whole negative sample z is calculated _r∈ z ₀feature wherein each negative sample z _rfeature be all d tie up x _r=(x _{r, 1}, x _{r, 2}..., x _{r, d}), solve the average of whole negative sample feature and variance ${\mathrm{DX}}_{0,c}=\frac{1}{n_0}\underset{r=1}{\overset{n_0}{\mathrm{\Σ}}}{x}_{0,r,c}^2-{\left({\mathrm{EX}}_{0,c}\right)}^2,$ Wherein c=1,2 ..., d;

I) all Weak Classifiers are upgraded

$\left\{\begin{array}{c}{\mathrm{\μ}}_{y,c}^{\′}\←(1-\mathrm{\λ}){\mathrm{\μ}}_{y,c}+\mathrm{\λ}{\mathrm{EX}}_{y,c}\\ {\mathrm{\σ}}_{y,c}^{\′}\←{[(1-\mathrm{\λ}){\mathrm{\σ}}_{y,c}^2+\mathrm{\λ}{\mathrm{DX}}_{y,c}+\mathrm{\λ}(1-\mathrm{\λ}){({\mathrm{\μ}}_{y,c}-{\mathrm{EX}}_{y,c})}^2]}^{1/2}\end{array}\right.$

Wherein c=1,2 ..., d.

Claims (3)

1. based on a dark situation video object method for real time tracking for textural characteristics, it is characterized in that, comprise initial phase and target tracking stage, described initial phase comprises the following steps:

1) when initialization, compute sparse sampling matrix Θ:

A) signal sampling matrix Φ is calculated;

B) compute sparse perception matrix Ψ;

C) compute sparse sampling matrix Θ, wherein Θ=Ψ Φ;

2) establishment one is by two classification Naive Bayes Classifier H (x) of 50 Bayes classifier cascades, each Weak Classifier h _c(x _c) be all based on representing two normal distributions of positive label y=1 and negative label y=0 wherein (μ _{y, c}, σ _{y, c}) represent that label is the parameter value of the Normal Discrimination curve of the Weak Classifier that the c dimensional feature of y is corresponding;

Described target tracking stage comprises the following steps:

1) target detection is carried out to the kth frame of video

A) with the target O that kth-1 frame traces into _k-1centered by carry out candidate samples collection, in kth frame, collect n _vindividual Euclidean distance meets $z_y=\left\{z\right|0\≤{\left|\right|z-O_{k-1}\left|\right|}^{l2}\≤r_y^{+}\}$ Candidate samples;

B) minimum rectangular area Uz (the z ∈ z comprising whole candidate samples is calculated _v), gray processing is carried out successively to this rectangular region image sheet, invariable rotary pattern ULBP encodes, vectorial integration, finally obtains vector product component I;

C) with the diagonal line of the nonzero element in sparse sampling matrix Θ for scale, from vector product component I, extract compressed encoding eigenwert z → x (z ∈ z of each candidate samples with diagonal line subtraction _v);

D) by the compressed encoding eigenwert x of each candidate samples _rthe Naive Bayes Classifier that input kth-1 frame trains, classifies to the feature of each candidate samples, and calculates classification score the x that classification score is maximum _rnamely r corresponding sample be the target O that kth frame traces into _k;

2) sorter of kth frame upgrades

A) with the target O that kth frame traces into _kcentered by carry out positive and negative sample collection, in kth frame, collect n ₁individual Euclidean distance meets z ₁=z|0≤|| z-O _k|| ^l2≤ r ₁ ⁺positive sample, in kth frame, collect n ₀individual Euclidean distance meets $z_0=\left\{z\right|r_0^{-}\≤{\left|\right|z-O_k\left|\right|}^{l2}\≤r_0^{+}\}$ Negative sample;

B) minimum rectangular area Uz (the z ∈ z comprising all positive negative samples is calculated ₁uz ₀), gray processing is carried out successively to this rectangular region image sheet, invariable rotary pattern ULBP encodes, vectorial integration, finally obtains vector product component I;

C) with the diagonal line of the nonzero element in sparse sampling matrix Θ for scale, from vector product component I, extract compressed encoding eigenwert z → x (z ∈ z of each positive negative sample with diagonal line subtraction ₁uz ₀);

D) sorter is upgraded

$\left\{\begin{array}{c}{\mathrm{\μ}}_{y,c}^{\′}\←(1-\mathrm{\λ}){\mathrm{\μ}}_{y,c}+\mathrm{\λ}{\mathrm{EX}}_{y,c}\\ {\mathrm{\σ}}_{y,c}^{\′}\←{[(1-\mathrm{\λ}){\mathrm{\σ}}_{y,c}^2+\mathrm{\λ}{\mathrm{DX}}_{y,c}+\mathrm{\λ}(1-\mathrm{\λ}){({\mathrm{\μ}}_{y,c}-{\mathrm{EX}}_{y,c})}^2]}^{1/2}\end{array}\right.$

Wherein with the average for positive sample (y=1) and negative sample (y=0) and variance respectively.

2. a kind of dark situation video object method for real time tracking based on textural characteristics according to claim 1, it is characterized in that, described target tracking stage is by gray-scale map I _grayto uniform pattern ULBP code pattern I _rULBPcoding method is:

1) pending pixel is as central pixel point, its gray-scale value g ₀with p the pixel g that distance is R _i∈ Γ _pthe difference of gray-scale value carry out binaryzation, and connect into the binary number of a p position, if the transition times U of 0 ~ 1 or 1 ~ 0 is not more than 2 in this binary number, then with the number of element 1, the uniform pattern ULBP as this pending pixel encodes, wherein $U=\underset{i=1}{\overset{p}{\mathrm{\Σ}}}1-\mathrm{\δ}[\mathrm{\ϵ}(g_{\mathrm{mod}(i+1,p)}-g_0)-\mathrm{\ϵ}(g_i-g_0)];$

3. a kind of dark situation video object method for real time tracking based on textural characteristics according to claim 1, it is characterized in that, described target tracking stage is by uniform pattern ULBP code pattern I _rULBPintegration method to vector product component I comprises the following steps:

1) statistics with histogram, each coding occurrence number of 9 dimension statistics with histogram of structure corresponding 0 ~ 8;

2) longitudinally cumulative, its step is

A) carry out flattening by row to H, after flattening, obtain a dimensional vector V _c;

B) to V _cadd up, the cumulative dimensional vector obtained meets

C) to V _{Σ C}carry out fractureing by row, obtain the equal-sized image with H, count H _i;

3) laterally cumulative, its step is

A) to H _iflatten by row, after flattening, obtain one dimension row vector V _r;

B) to V _radd up, the cumulative dimensional vector obtained meets

C) to V _{Σ R}carry out fractureing by row, obtain the equal-sized image with H, count H _ii;

H _iibe the image I after process, in H, the statistic histogram of anyon matrix all can by H _iicarry out diagonal line subtraction to try to achieve.