CN102930558A - Real-time tracking method for infrared image target with multi-feature fusion - Google Patents

Abstract

The invention discloses a multi-feature fusion infrared image target real-time tracking method, comprising the following steps: initializing the target tracking point position; initializing the target model; calculating the target candidate model; calculating the joint feature Bhattacharyya coefficient and the weight coefficient between features; The frame target tracks the new position; estimates the joint feature Bhattacharyya coefficient of the new position; compares the two joint feature Bhattacharyya coefficients before and after, and outputs the result. The invention self-adaptively calculates the weight coefficient among multiple features, enhances the robustness of target tracking, ensures the stability of target tracking, solves the problem of tracking point drift caused by the instability of a single feature, and effectively improves the target tracking accuracy.

Description

A kind of infrared image object real-time tracking method of many Fusion Features

Technical field

The present invention designs a kind of infrared image method for tracking target of many Fusion Features, particularly a kind of infrared image method for tracking target of suitable hardware real-time implementation.

Background technology

In recent years, along with the development of integrated circuit technology and infra-red material, infrared imagery technique has been obtained very much progress, is used widely in national defense construction and national economy field.Yet, to compare with visible images, the infrared image signal to noise ratio (S/N ratio) is relatively low, therefore can only provide limited information when carrying out the infrared image target detection and following the tracks of.Because target signature is not obvious in the infrared image, has the large problems such as background clutter, causes the accurate tracking of infrared image target to become more difficult.

At present, target tracking algorism is divided into based on the tracking of model with based on the tracking two large classes of outward appearance.Compare with the model following method, the outward appearance tracing has avoided setting up the complex process of model, has wider engineering practical value.Wherein, average drifting track algorithm, robust simple because of it, characteristics that real-time is good are used widely in target following.Average drifting is a kind of without ginseng density calculation method, by the iterative search distribution pattern the most similar to sample distribution repeatedly.The people such as Comaniciu propose a kind of average drifting target tracking algorism by seeking the maximum value of color of object histogram and candidate target color histogram similarity.The people such as Chu are used for the Kalman wave filter primary iteration position of prediction Mean Shift, but when target is seriously blocked, because the source location that Mean Shift algorithm searches out is inaccurate, have certain deviation.A kind of adaptive tracking method that can choose easy identification color characteristic of the propositions such as Collins, wherein the candidate color feature set comprises 49 stack features that calculated by pixel R, G, the linear combination of B value. because the Candidate Set that adopts is larger, the computing expense of Feature Selection is also very large.Therefore, there is following shortcoming in existing target tracking algorism: (1) classical target tracking algorism adopts single features to describe target, poor anti jamming capability; (2) most existing many features target tracking algorisms only utilize the weight coefficient between the present frame calculated characteristics, carry out complexity when target and change the track algorithm poor robustness; (3) most existing track algorithms are in the situation that non-rigid deformation, partial occlusion and overlapping appear in target, and tracking accuracy descends, even track rejection occurs; (4) most existing track algorithms have increased algorithm complex greatly when improving target tracking accuracy, are difficult for the hardware real-time implementation.

Summary of the invention

Goal of the invention: technical matters to be solved by this invention is for the deficiencies in the prior art, and a kind of infrared image object real-time tracking method of many Fusion Features is provided.

In order to solve the problems of the technologies described above, the invention discloses a kind of infrared image object real-time tracking method of many Fusion Features, may further comprise the steps:

(1) initialization tracking position position y ₀Initial trace point is by artificial appointment;

(2) initialization object module is with initial trace point y ₀Centered by set up target gray level model q ₁With target LBP texture model q ₂(local binary pattern, LBP) local binary patterns.

(3) calculate the target candidate model, according to the trace point position y of target ₀, calculated candidate target gray level model p ₁(y ₀) and candidate target LBP texture model p ₂(y ₀);

(4) utilize gray feature Bhattacharyya(Batachelia Bhattacharyya, referring to Visual C++ Digital Image Processing, the 466th page, author: Xie Fengying, first published in 2008, Electronic Industry Press.) coefficient ρ ₁Bhattacharyya coefficient ρ with the LBP textural characteristics ₂, and the weight coefficient α of gray feature ₁Weight coefficient α with the LBP textural characteristics ₂, calculating location y ₀The union feature Bhattacharyya of place coefficient ρ, expression formula is as follows:

ρ＝α ₁·ρ ₁+α ₂·ρ ₂；

(5) calculate present frame target reposition y ₁

(6) utilize gray feature Bhattacharyya coefficient ρ ' ₁With LBP textural characteristics Bhattacharyya coefficient ρ ' ₂, and the weight coefficient α ' of gray feature ₁Weight coefficient α ' with the LBP textural characteristics ₂, calculating location y ₁The union feature Bhattacharyya of place coefficient ρ ', expression formula is as follows;

ρ＝α′ ₁·ρ′ ₁+α′ ₂·ρ′ ₂，

(7) when ρ '＜ρ,

Otherwise y ₁Remain unchanged;

(8) if | (y ₀-y ₁) |＜ε, stop to calculate, otherwise, with y ₁Assignment is to y ₀And execution in step (3), wherein, ε is the error constant coefficient.

In the step (2), target gray level model q ₁For:

$q_1={\left\{q_u\right\}}_{u=1...m_1},$ $\underset{u=1}{\overset{m_1}{\mathrm{\Σ}}}{q}_u=1,$

Target LBP texture model q ₂For:

$q_2={\left\{q_v\right\}}_{v=1...m_2},$ $\underset{v=1}{\overset{m_2}{\mathrm{\Σ}}}{q}_v=1,$

Wherein, q _uBe the probability density at different levels of target gray level model gray feature, q _vBe the probability density at different levels of LBP textural characteristics, m ₁Be the maximum quantification progression scope of target gray level model gray feature, m ₂Be the maximum quantification progression scope of LBP textural characteristics, u represents grey level quantization progression, and v represents that texture quantizes progression.

In the step (3), candidate target gray level model p ₁For:

$p_1={\left\{p_u\right\}}_{u=1...m_1},$ $\underset{u=1}{\overset{m_1}{\mathrm{\Σ}}}{p}_u=1,$

Candidate target LBP texture model p ₂For:

$p_2={\left\{p_v\right\}}_{v=1...m_2},$ $\underset{v=1}{\overset{m_2}{\mathrm{\Σ}}}{p}_v=1,$

p _uBe the probability density at different levels of target gray level model gray feature, p _vBe the probability density at different levels of target gray level model gray feature and LBP textural characteristics, m ₁Be the maximum quantification progression scope of target gray level model gray feature, m ₂Be the maximum quantification progression scope of LBP textural characteristics, u represents grey level quantization progression, and v represents that texture quantizes progression.

The weight coefficient α of gray feature in the step (4) ₁Weight coefficient α with the LBP textural characteristics ₂, the weight coefficient α ' of gray feature in the step (6) ₁Weight coefficient α ' with the LBP textural characteristics ₂The employing iterative manner is upgraded, and calculating formula is as follows respectively:

α ₁=(1-λ)·α _1,old+λ·α _1,cur，

α ₂=(1-λ)·α _2,old+λ·α _2,cur，

α′ ₁=(1-λ)·α′ _1，old+λ·α′ _1,cur，

α′ ₂=(1-λ)·α′ _2,old+λ·α′ _2,cur，

Wherein, α _{1, old}And α _{2, old}Respectively the weight coefficient of the middle previous frame gray feature of step (4) and LBP textural characteristics, α _{1, cur}And α _{2, cur}Respectively the weight coefficient of the middle present frame gray feature of step (4) and LBP textural characteristics, α ' _{1, old}And α ' _{2, old}Respectively the weight coefficient of the middle previous frame gray feature of step (6) and LBP textural characteristics, α ' _{1, cur}And α ' _{2, cur}Be the weight coefficient of the middle present frame gray feature of step (6) and LBP textural characteristics, λ is scale-up factor.0≤λ≤1, the speed of convergence of decision weight coefficient, λ value is larger, and speed of convergence is faster, and tracking maneuverability is stronger, and λ value is less, and speed of convergence is slower, and tracking stability is better.

The weight coefficient α of present frame gray feature in the step (4) _{1, cur}Weight coefficient α with the LBP textural characteristics _{2, cur}, the weight coefficient α ' of present frame gray feature in the step (6) _{1, cur}Weight coefficient α ' with the LBP textural characteristics _{2, cur}, calculating formula is as follows respectively:

${\mathrm{\α}}_{1,\mathrm{cur}}=\frac{{\mathrm{\ρ}}_1}{\sqrt{{\mathrm{\ρ}}_1^2+{\mathrm{\ρ}}_2^2}},$

${\mathrm{\α}}_{2,\mathrm{cur}}=\frac{{\mathrm{\ρ}}_2}{\sqrt{{\mathrm{\ρ}}_1^2+{\mathrm{\ρ}}_2^2}},$

${\mathrm{\α}}_{1,\mathrm{cur}}^{\′}=\frac{{\mathrm{\ρ}}_1^{\′}}{\sqrt{{\mathrm{\ρ}}_1^{\′}\·{\mathrm{\ρ}}_1^{\′}+{\mathrm{\ρ}}_2^{\′}\·{\mathrm{\ρ}}_2^{\′}}},$

${\mathrm{\α}}_{2,\mathrm{cur}}^{\′}=\frac{{\mathrm{\ρ}}_2^{\′}}{\sqrt{{\mathrm{\ρ}}_1^{\′}\·{\mathrm{\ρ}}_1^{\′}+{\mathrm{\ρ}}_2^{\′}\·{\mathrm{\ρ}}_2^{\′}}},$

Wherein, ρ ₁The Bhattacharyya coefficient of gray feature in the step (4), ρ ₂The Bhattacharyya coefficient of LBP textural characteristics in the step (4), ρ ' ₁The Bhattacharyya coefficient of gray feature in the step (6), ρ ' ₂It is the Bhattacharyya coefficient of LBP textural characteristics in the step (6).

Gray feature Bhattacharyya coefficient ρ ' in the step (6) ₁Bhattacharyya coefficient ρ ' with the LBP textural characteristics ₂, and the weight coefficient α ' of gray feature ₁Weight coefficient α ' with the LBP textural characteristics ₂Obtain by following formula:

${\mathrm{\ρ}}_1^{\′}=\underset{u=1}{\overset{m_1}{\mathrm{\Σ}}}\sqrt{p_u^{\′}\·q_u},$

${\mathrm{\ρ}}_2^{\′}=\underset{u=1}{\overset{m_2}{\mathrm{\Σ}}}\sqrt{p_v^{\′}\·q_v},$

P ' _uPosition y ₁The probability density at different levels of the target gray level model gray feature at place, p ' _vPosition y ₁The probability density at different levels of the LBP of place textural characteristics;

${\mathrm{\α}}_1^{\′}=(1-\mathrm{\λ})\·{\mathrm{\α}}_{1,\mathrm{old}}^{\′}+\mathrm{\λ}\·\frac{{\mathrm{\ρ}}_1^{\′}}{\sqrt{{\mathrm{\ρ}}_1^{\′}\·{\mathrm{\ρ}}_1^{\′}+{\mathrm{\ρ}}_2^{\′}\·{\mathrm{\ρ}}_2^{\′}}}$

${\mathrm{\α}}_2^{\′}=(1-\mathrm{\λ})\·{\mathrm{\α}}_{2,\mathrm{old}}^{\′}+\mathrm{\λ}\·\frac{{\mathrm{\ρ}}_2^{\′}}{\sqrt{{\mathrm{\ρ}}_1^{\′}\·{\mathrm{\ρ}}_1^{\′}+{\mathrm{\ρ}}_2^{\′}\·{\mathrm{\ρ}}_2^{\′}}},$

Wherein, α ' _{1, old}The weight coefficient of previous frame gray feature, α ' _{2, old}Be the weight coefficient of previous frame LBP textural characteristics, λ is scale-up factor.0≤λ≤1, the speed of convergence of decision weight coefficient, λ value is larger, and speed of convergence is faster, and tracking maneuverability is stronger, and λ value is less, and speed of convergence is slower, and tracking stability is better.

In the infrared image object real-time tracking method of the many Fusion Features of the present invention, use Yi Pannieqiekefu (Epanechnikov) kernel function to calculate gray feature probability histogram and LBP textural characteristics probability histogram.

The present invention compared with prior art has following remarkable advantage: (1) according to characteristic remarkable and the similarity of target and background, self-adaptation is calculated the weight coefficient between many features, has strengthened the target following robustness; (2) the employing iterative manner is upgraded the weight coefficient between many features, has guaranteed target following stability; (3) utilize many Fusion Features method for tracking target that the infrared image target is followed the tracks of, solved the unstable trace point drifting problem that causes of single features, Effective Raise target tracking accuracy; (4) many Fusion Features method for tracking target of the present invention's proposition does not exist high exponent arithmetic(al) and labyrinth, and algorithm operation quantity is little, is easy to the hardware real-time implementation.

Description of drawings

Below in conjunction with the drawings and specific embodiments the present invention is done further to specify, the advantage of above and other of the present invention aspect will become apparent.

Fig. 1 is process flow diagram of the present invention.

Fig. 2 a ~ 2d is traditional single feature (gray scale) infrared image target following result.

Fig. 3 a ~ 3d is the infrared image target following result of the many Fusion Features of the present invention.

Embodiment

In the infrared image object real-time tracking method of the many Fusion Features of the present invention, utilize gray feature and LBP textural characteristics to describe the infrared image clarification of objective.

Eight neighborhood LBP textural characteristics expression formula LBP _8,1As follows:

$\mathrm{LB}{P}_{\mathrm{8,1}}=\underset{n=0}{\overset{7}{\mathrm{\Σ}}}s(g_n-g_c)\·2^n,$

$s\left(x\right)=\left\{\begin{array}{cc}1,& x\&GreaterEqual;0\\ 0,& x<0\end{array}\right.$

Wherein, g _cCurrent point, g _nNeighborhood point on every side, n=0..7.

In the infrared image object real-time tracking method of the many Fusion Features of the present invention, utilize the Bhattacharyya coefficient to describe similarity between object module and target candidate model.

Bhattacharyya coefficient ρ _BhaExpression formula is as follows:

${\mathrm{\&rho;}}_{\mathrm{Bha}}=\underset{u=1}{\overset{m}{\mathrm{\&Sigma;}}}\sqrt{p\&CenterDot;q},$

Wherein, p is the target candidate model, and q is object module.

In the infrared image object real-time tracking method of the many Fusion Features of the present invention, the weight coefficient α of gray feature ₁Weight coefficient α with the LBP textural characteristics ₂The employing iterative manner is upgraded, and expression formula is as follows:

α ₁=(1-λ)·α _1,old+λ·α _1，cur

α ₂=(1-λ)·α _2,old+λ·α _2,cur

Wherein, α _{1, old}And α _{2, old}Respectively the weight coefficient of previous frame gray feature and LBP textural characteristics, α _{1, cur}And α _{2, cur}Be the weight coefficient of present frame gray feature and LBP textural characteristics, λ is scale-up factor.

In the infrared image object real-time tracking method of the many Fusion Features of the present invention, the weight coefficient α of present frame gray feature _{1, cur}Weight coefficient α with the LBP textural characteristics _{2, cur}Expression formula is as follows respectively:

${\mathrm{\&alpha;}}_{1,\mathrm{cur}}=\frac{{\mathrm{\&rho;}}_1}{\sqrt{{\mathrm{\&rho;}}_1^2+{\mathrm{\&rho;}}_2^2}},$

${\mathrm{\&alpha;}}_{2,\mathrm{cur}}=\frac{{\mathrm{\&rho;}}_2}{\sqrt{{\mathrm{\&rho;}}_1^2+{\mathrm{\&rho;}}_2^2}},$

Wherein, ρ ₁The Bhattacharyya coefficient of gray feature, ρ ₂It is the Bhattacharyya coefficient of LBP textural characteristics.

Embodiment 1

As shown in Figure 1, the below illustrates the infrared image object real-time tracking method of the many Fusion Features of the present invention with example.The number of pixels 320 * 240 of infrared image, frame frequency 25HZ.The thermal infrared imager imaging is by the special image disposable plates of optical fiber transmission to the DSP+FPGA framework, and the infrared image target following of many Fusion Features realizes in dsp processor, satisfies the demand of processing in real time, and the implementation step is as follows:

(1) initialization tracking position position y ₀, initial trace point is by artificial appointment.

The artificial initial target trace point position (i, j) of specifying, i=80, j=100(is shown in Figure 2), set Yi Pannieqiekefu (Epanechnikov) kernel function bandwidth h=10.

(2) initialization object module is set up target gray level model q according to gray feature ₁, the LBP textural characteristics of calculating target is set up target texture model q in conjunction with the LBP textural characteristics ₂

Calculating is centered by initial trace point position (80,100), and bandwidth h=10 is the image I of scope _LBPThe LBP textural characteristics, expression formula is as follows:

$I_{\mathrm{LBP}}=\underset{k1=75}{\overset{85}{\mathrm{\&Sigma;}}}\underset{k2=95}{\overset{105}{\mathrm{\&Sigma;}}}\mathrm{LB}{P}_{\mathrm{8,1}}(k1,k2)$

Eight neighborhood LBP textural characteristics expression formula LBP _8,1As follows:

$\mathrm{LB}{P}_{\mathrm{8,1}}(k1,k2)=\underset{n=0}{\overset{7}{\mathrm{\&Sigma;}}}s(g_n-g_c)\&CenterDot;2^n,$

$s\left(x\right)=\left\{\begin{array}{cc}1,& x\&GreaterEqual;0\\ 0,& x<0\end{array}\right.,$

Wherein, g _cThe current goal point, c=k2*320+k1, g _nG _cNeighborhood point on every side, n=0..7.

Target gray level model q ₁For:

$q_1={\left\{q_u\right\}}_{u=1...m_1},$ $\underset{u=1}{\overset{m_1}{\mathrm{\&Sigma;}}}{q}_u=1,$

$q_u=C\underset{i=1}{\overset{n}{\mathrm{\&Sigma;}}}k\left[{\left|\right|\frac{y_0-x_i}{h}\left|\right|}^2\right]\&CenterDot;\mathrm{\&delta;}[b_1\left(x_i\right)-\mathrm{\&mu;}],$

Target LBP texture model q ₂For:

$q_2={\left\{q_v\right\}}_{v=1...m_2},$ $\underset{v=1}{\overset{m_2}{\mathrm{\&Sigma;}}}{q}_v=1,$

$q_v=C\underset{i=1}{\overset{n^{\&prime;}}{\mathrm{\&Sigma;}}}k\left[{\left|\right|\frac{y_0-x_i}{h}\left|\right|}^2\right]\&CenterDot;\mathrm{\&delta;}[b_2\left(x_i\right)-v],$

Wherein, q _uAnd q _vThe probability density at different levels that represent respectively object module gray feature and LBP textural characteristics, m ₁=255 and m ₂=255 represent respectively the quantification progression of object module gray feature and LBP textural characteristics, function b ₁() is to be positioned at x _iPixel to the reflection of gray feature index, function b ₂() is to be positioned at x _iPixel to the reflection of LBP textural characteristics index, δ is the Delta function, C is normalization coefficient, μ=1...255, v=1...255.

(3) calculate the target candidate model.According to trace point position y ₀, calculated candidate target gray level model p ₁(y ₀) and target texture candidate family p ₂(y ₀);

Candidate target gray level model p ₁For:

$p_1={\left\{p_u\right\}}_{u=1...m_1},$ $\underset{u=1}{\overset{m_1}{\mathrm{\&Sigma;}}}{p}_u=1,$

$p_u=C\underset{i=1}{\overset{n}{\mathrm{\&Sigma;}}}k\left[{\left|\right|\frac{y_0-x_i}{h}\left|\right|}^2\right]\&CenterDot;\mathrm{\&delta;}[b_1\left(x_i\right)-\mathrm{\&mu;}],$

Candidate target LBP texture model p ₂For:

$p_2={\left\{p_v\right\}}_{v=1...m_2},$ $\underset{v=1}{\overset{m_2}{\mathrm{\&Sigma;}}}{p}_v=1,$

$p_v=C\underset{i=1}{\overset{n^{\&prime;}}{\mathrm{\&Sigma;}}}k\left[{\left|\right|\frac{y_0-x_i}{h}\left|\right|}^2\right]\&CenterDot;\mathrm{\&delta;}[b_2\left(x_i\right)-v],$

p _uAnd p _vThe probability density at different levels that represent respectively object module gray feature and LBP textural characteristics, m ₁=255 and m ₂=255 represent respectively the quantification progression of object module gray feature and LBP textural characteristics, function b ₁() is to be positioned at x _iPixel to the reflection of gray feature index, function b ₂() is to be positioned at x _iPixel to the reflection of LBP textural characteristics index, δ is the Delta function, C is normalization coefficient, μ=1...255, v=1...255.

(4) calculate respectively the Bhattacharyya coefficient ρ of gray feature and LBP textural characteristics ₁, ρ ₂And weight coefficient α ₁, α ₂, utilize ρ ₁, α ₁, ρ ₂, α ₂Calculating location y ₀The union feature Bhattacharyya coefficient ρ at place;

Union feature Bhattacharyya coefficient ρ is described as:

ρ=α ₁·ρ ₁+α ₂·ρ ₂，

${\mathrm{\&rho;}}_1=\underset{u=1}{\overset{m_1}{\mathrm{\&Sigma;}}}\sqrt{p_u\&CenterDot;q_u},$

${\mathrm{\&rho;}}_2=\underset{u=1}{\overset{m_2}{\mathrm{\&Sigma;}}}\sqrt{p_v\&CenterDot;q_v},$

Gray feature weight coefficient α ₁, LBP textural characteristics α ₂The renewal expression formula is as follows:

${\mathrm{\&alpha;}}_1=(1-\mathrm{\&lambda;})\&CenterDot;{\mathrm{\&alpha;}}_{1,\mathrm{old}}+\mathrm{\&lambda;}\&CenterDot;{\mathrm{\&alpha;}}_{1,\mathrm{cur}}=(1-\mathrm{\&lambda;})\&CenterDot;{\mathrm{\&alpha;}}_{1,\mathrm{old}}+\mathrm{\&lambda;}\&CenterDot;\frac{{\mathrm{\&rho;}}_1}{\sqrt{{\mathrm{\&rho;}}_1^2+{\mathrm{\&rho;}}_2^2}},$

${\mathrm{\&alpha;}}_2=(1-\mathrm{\&lambda;})\&CenterDot;{\mathrm{\&alpha;}}_{2,\mathrm{old}}+\mathrm{\&lambda;}\&CenterDot;{\mathrm{\&alpha;}}_{2,\mathrm{cur}}=(1-\mathrm{\&lambda;})\&CenterDot;{\mathrm{\&alpha;}}_{2,\mathrm{old}}+\mathrm{\&lambda;}\&CenterDot;\frac{{\mathrm{\&rho;}}_2}{\sqrt{{\mathrm{\&rho;}}_1^2+{\mathrm{\&rho;}}_2^2}},$

Wherein, α _{1, old}And α _{2, old}The previous frame weight coefficient, α _{1, cur}And α _{2, cur}Be the present frame weight coefficient, λ is scale-up factor.

(5) calculate present frame target following reposition y ₁

$y_1={\mathrm{\&alpha;}}_1\&CenterDot;\frac{\underset{i=1}{\overset{n^{\&prime;}}{\mathrm{\&Sigma;}}}{x}_i\&CenterDot;w_{i,1}}{\underset{i=1}{\overset{n^{\&prime;}}{\mathrm{\&Sigma;}}}{w}_{i,1}}+{\mathrm{\&alpha;}}_2\&CenterDot;\frac{\underset{i=1}{\overset{n^{\&prime;}}{\mathrm{\&Sigma;}}}{x}_i\&CenterDot;w_{i,2}}{\underset{i=1}{\overset{n^{\&prime;}}{\mathrm{\&Sigma;}}}{w}_{i,2}},$

$w_{i,1}=\underset{u=1}{\overset{m_1}{\mathrm{\&Sigma;}}}\sqrt{\frac{q_u}{p_u\left(y_0\right)}}\&CenterDot;\mathrm{\&delta;}[b_1\left(x_i\right)-u],$

$w_{i,2}=\underset{v=1}{\overset{m_2}{\mathrm{\&Sigma;}}}\sqrt{\frac{q_v}{p_v\left(y_0\right)}}\&CenterDot;\mathrm{\&delta;}[b_2\left(x_i\right)-v],$

Wherein, n' is candidate target pixel number, and h is the kernel function bandwidth, q _u, q _v, p _u, p _v, α ₁, α ₂, x _i, b ₁(), b ₂The implication of () is identical with the definition in step (2), (3), (4).

(6) utilize the Bhattacharyya coefficient ρ ' of gray feature and LBP textural characteristics ₁, ρ ' ₂And weight coefficient α ' ₁, α ' ₂, calculating location y ₁The union feature Bhattacharyya of place coefficient ρ ', expression formula is as follows;

ρ'=α′ ₁·ρ′ ₁+α′ ₂·ρ′ ₂，

${\mathrm{\&rho;}}_1^{\&prime;}=\underset{u=1}{\overset{m_1}{\mathrm{\&Sigma;}}}\sqrt{p_u^{\&prime;}\&CenterDot;q_u},$

${\mathrm{\&rho;}}_2^{\&prime;}=\underset{u=1}{\overset{m_2}{\mathrm{\&Sigma;}}}\sqrt{p_v^{\&prime;}\&CenterDot;q_v},$

Gray feature weight coefficient α ' ₁, LBP textural characteristics α ' ₂The renewal expression formula is as follows:

${\mathrm{\&alpha;}}_1^{\&prime;}=(1-\mathrm{\&lambda;})\&CenterDot;{\mathrm{\&alpha;}}_{1,\mathrm{old}}^{\&prime;}+\mathrm{\&lambda;}\&CenterDot;\frac{{\mathrm{\&rho;}}_1^{\&prime;}}{\sqrt{{\mathrm{\&rho;}}_1^{\&prime;}\&CenterDot;{\mathrm{\&rho;}}_1^{\&prime;}+{\mathrm{\&rho;}}_2^{\&prime;}\&CenterDot;{\mathrm{\&rho;}}_2^{\&prime;}}}$

${\mathrm{\&alpha;}}_2^{\&prime;}=(1-\mathrm{\&lambda;})\&CenterDot;{\mathrm{\&alpha;}}_{2,\mathrm{old}}^{\&prime;}+\mathrm{\&lambda;}\&CenterDot;\frac{{\mathrm{\&rho;}}_2^{\&prime;}}{\sqrt{{\mathrm{\&rho;}}_1^{\&prime;}\&CenterDot;{\mathrm{\&rho;}}_1^{\&prime;}+{\mathrm{\&rho;}}_2^{\&prime;}\&CenterDot;{\mathrm{\&rho;}}_2^{\&prime;}}},$

Wherein, α ' _{1, old}And α ' _{2, old}Be the previous frame weight coefficient, λ is scale-up factor.

(7) when ρ '＜ρ, Otherwise y ₁Remain unchanged;

(8) if abs is (y ₀-y ₁)＜0.01 item stops, otherwise, y ₀← y ₁, execution in step (3).

Fig. 2 is conventional art, and Fig. 3 is for the single feature (gray feature) of only utilization that obtains according to the present embodiment and the infrared image target following result of many Fusion Features, because be infrared image, so greyscale color unavoidably occurs.Wherein Fig. 2 a, 2b, 2c, 2d represent respectively the image of the 20th frame, the 80th frame, the 140th frame and the 200th frame, and Fig. 3 a, 3b, 3c, 3d represent respectively the image of the 20th frame, the 80th frame, the 140th frame and the 200th frame.Comparison diagram 2 and Fig. 3 find: only utilize single feature that the target following meeting is caused that tracing process is unstable, tracking accuracy is poor, swings at random such as Fig. 2 tracking gate; Utilize the tracking of many Fusion Features can Effective Raise tracking accuracy, such as Fig. 3 tracking gate all the time near the object type heart.

The invention provides a kind of infrared image object real-time tracking method of many Fusion Features; method and the approach of this technical scheme of specific implementation are a lot; the above only is preferred implementation of the present invention; should be understood that; for those skilled in the art; under the prerequisite that does not break away from the principle of the invention, can also make some improvements and modifications, these improvements and modifications also should be considered as protection scope of the present invention.In the present embodiment not clear and definite each ingredient all available prior art realized.

Claims (6)

1. the infrared image object real-time tracking method of Fusion Features more than a kind is characterized in that, may further comprise the steps:

(1) gives the initial trace point position y that sets the goal ₀

(2) initialization object module is with initial trace point y ₀Centered by set up target gray level model q ₁With target LBP texture model q ₂

(3) calculate the target candidate model, according to the trace point position y of target ₀, calculated candidate target gray level model p ₁(y ₀) and candidate target LBP texture model p ₂(y ₀);

(4) utilize gray feature Bhattacharyya coefficient ρ ₁Bhattacharyya coefficient ρ with the LBP textural characteristics ₂, and the weight coefficient α of gray feature ₁Weight coefficient α with the LBP textural characteristics ₂, calculating location y ₀The union feature Bhattacharyya of place coefficient ρ, expression formula is as follows:

ρ＝α ₁·ρ ₁+α ₂·ρ ₂；

(5) calculate present frame target reposition y ₁

(6) utilize gray feature Bhattacharyya coefficient ρ ' ₁With LBP textural characteristics Bhattacharyya coefficient ρ ' ₂, and the weight coefficient α ' of gray feature ₁Weight coefficient α ' with the LBP textural characteristics ₂, calculating location y ₁The union feature Bhattacharyya of place coefficient ρ ', expression formula is as follows;

ρ＝α′ ₁·ρ′ ₁+α′ ₂·ρ′ ₂，

(7) when ρ '＜ρ,

Otherwise y ₁Remain unchanged;

(8) if | (y ₀-y ₁) |＜ε, stop to calculate, otherwise, with y ₁Assignment is to y ₀And execution in step (3), wherein, ε is the error constant coefficient.

2. the infrared image object real-time tracking method of a kind of many Fusion Features according to claim 1 is characterized in that, in the step (2), and target gray level model q ₁For:

$q_1={\left\{q_u\right\}}_{u=1...m_1},$ $\underset{u=1}{\overset{m_1}{\mathrm{\&Sigma;}}}{q}_u=1,$

Target LBP texture model q ₂For:

$q_2={\left\{q_v\right\}}_{v=1...m_2},$ $\underset{v=1}{\overset{m_2}{\mathrm{\&Sigma;}}}{q}_v=1,$

Wherein, q _uBe the probability density at different levels of target gray level model gray feature, q _vBe the probability density at different levels of LBP textural characteristics, m ₁Be the maximum quantification progression scope of target gray level model gray feature, m ₂Be the maximum quantification progression scope of LBP textural characteristics, u represents grey level quantization progression, and v represents that texture quantizes progression.

3. the infrared image object real-time tracking method of a kind of many Fusion Features according to claim 1 is characterized in that, in the step (3), and candidate target gray level model p ₁For:

$p_1={\left\{p_u\right\}}_{u=1...m_1},$ $\underset{u=1}{\overset{m_1}{\mathrm{\&Sigma;}}}{p}_u=1,$

Candidate target LBP texture model p ₂For:

$p_2={\left\{p_v\right\}}_{v=1...m_2},$ $\underset{v=1}{\overset{m_2}{\mathrm{\&Sigma;}}}{p}_v=1,$

p _uBe the probability density at different levels of target gray level model gray feature, p _vBe the probability density at different levels of target gray level model gray feature and LBP textural characteristics, m ₁Be the maximum quantification progression scope of target gray level model gray feature, m ₂Be the maximum quantification progression scope of LBP textural characteristics, u represents grey level quantization progression, and v represents that texture quantizes progression.

4. the infrared image object real-time tracking method of a kind of many Fusion Features according to claim 1 is characterized in that, the weight coefficient α of gray feature in the step (4) ₁Weight coefficient α with the LBP textural characteristics ₂, the weight coefficient α ' 1 of gray feature and the weight coefficient α ' of LBP textural characteristics in the step (6) ₂The employing iterative manner is upgraded, and calculating formula is as follows respectively:

α ₁=(1-λ)·α _1,old+λ·α _1,cur，

α ₂=(1-λ)·α _2,old+λ·α _2,cur，

α′ ₁=(1-λ)·α′ _1，old+λ·α′ _1,cur，

α′ ₂=(1-λ)·α′ _2,old+λ·α′ _2,cur，

Wherein, α _{1, old}And α _{2, old}Respectively the weight coefficient of the middle previous frame gray feature of step (4) and LBP textural characteristics, α _{1, cur}And α _{2, cur}Respectively the weight coefficient of the middle present frame gray feature of step (4) and LBP textural characteristics, α ' _{1, old}And α ' _{2, old}Respectively the weight coefficient of the middle previous frame gray feature of step (6) and LBP textural characteristics, α ' _{1, cur}And α ' _{2, cur}Be the weight coefficient of the middle present frame gray feature of step (6) and LBP textural characteristics, λ is scale-up factor.

5. the infrared image object real-time tracking method of a kind of many Fusion Features according to claim 4 is characterized in that, the weight coefficient α of present frame gray feature in the step (4) _{1, cur}Weight coefficient α with the LBP textural characteristics _{2, cur}, the weight coefficient α ' of present frame gray feature in the step (6) _{1, cur}Weight coefficient α ' with the LBP textural characteristics _{2, cur}, calculating formula is as follows respectively:

${\mathrm{\&alpha;}}_{1,\mathrm{cur}}=\frac{{\mathrm{\&rho;}}_1}{\sqrt{{\mathrm{\&rho;}}_1^2+{\mathrm{\&rho;}}_2^2}},$

${\mathrm{\&alpha;}}_{2,\mathrm{cur}}=\frac{{\mathrm{\&rho;}}_2}{\sqrt{{\mathrm{\&rho;}}_1^2+{\mathrm{\&rho;}}_2^2}},$

${\mathrm{\&alpha;}}_{1,\mathrm{cur}}^{\&prime;}=\frac{{\mathrm{\&rho;}}_1^{\&prime;}}{\sqrt{{\mathrm{\&rho;}}_1^{\&prime;}\&CenterDot;{\mathrm{\&rho;}}_1^{\&prime;}+{\mathrm{\&rho;}}_2^{\&prime;}\&CenterDot;{\mathrm{\&rho;}}_2^{\&prime;}}},$

${\mathrm{\&alpha;}}_{2,\mathrm{cur}}^{\&prime;}=\frac{{\mathrm{\&rho;}}_2^{\&prime;}}{\sqrt{{\mathrm{\&rho;}}_1^{\&prime;}\&CenterDot;{\mathrm{\&rho;}}_1^{\&prime;}+{\mathrm{\&rho;}}_2^{\&prime;}\&CenterDot;{\mathrm{\&rho;}}_2^{\&prime;}}},$

Wherein, ρ ₁The Bhattacharyya coefficient of gray feature in the step (4), ρ ₂The Bhattacharyya coefficient of LBP textural characteristics in the step (4), ρ ' ₁The Bhattacharyya coefficient of gray feature in the step (6), ρ ' ₂It is the Bhattacharyya coefficient of LBP textural characteristics in the step (6).

6. the infrared image object real-time tracking method of a kind of many Fusion Features according to claim 5 is characterized in that, gray feature Bhattacharyya coefficient ρ ' in the step (6) ₁Bhattacharyya coefficient ρ ' with the LBP textural characteristics ₂, and the weight coefficient α ' of gray feature ₁Weight coefficient α ' with the LBP textural characteristics ₂Obtain by following formula:

${\mathrm{\&rho;}}_1^{\&prime;}=\underset{u=1}{\overset{m_1}{\mathrm{\&Sigma;}}}\sqrt{p_u^{\&prime;}\&CenterDot;q_u},$

${\mathrm{\&rho;}}_2^{\&prime;}=\underset{u=1}{\overset{m_2}{\mathrm{\&Sigma;}}}\sqrt{p_v^{\&prime;}\&CenterDot;q_v},$

P ' _uPosition y ₁The probability density at different levels of the target gray level model gray feature at place, p ' _vPosition y ₁The probability density at different levels of the LBP of place textural characteristics;

${\mathrm{\&alpha;}}_1^{\&prime;}=(1-\mathrm{\&lambda;})\&CenterDot;{\mathrm{\&alpha;}}_{1,\mathrm{old}}^{\&prime;}+\mathrm{\&lambda;}\&CenterDot;\frac{{\mathrm{\&rho;}}_1^{\&prime;}}{\sqrt{{\mathrm{\&rho;}}_1^{\&prime;}\&CenterDot;{\mathrm{\&rho;}}_1^{\&prime;}+{\mathrm{\&rho;}}_2^{\&prime;}\&CenterDot;{\mathrm{\&rho;}}_2^{\&prime;}}}$

${\mathrm{\&alpha;}}_2^{\&prime;}=(1-\mathrm{\&lambda;})\&CenterDot;{\mathrm{\&alpha;}}_{2,\mathrm{old}}^{\&prime;}+\mathrm{\&lambda;}\&CenterDot;\frac{{\mathrm{\&rho;}}_2^{\&prime;}}{\sqrt{{\mathrm{\&rho;}}_1^{\&prime;}\&CenterDot;{\mathrm{\&rho;}}_1^{\&prime;}+{\mathrm{\&rho;}}_2^{\&prime;}\&CenterDot;{\mathrm{\&rho;}}_2^{\&prime;}}},$

Wherein, α ' _{1, old}The weight coefficient of previous frame gray feature, α ' _{2, old}Be the weight coefficient of previous frame LBP textural characteristics, λ is scale-up factor.

Images