CN105320443A - Method and device for switching channel through gesture - Google Patents

Abstract

The invention discloses a method for switching a channel through a gesture. The method comprises the steps that an image of the gesture is collected; a Laplacian matrix is structured and subjected to eigen-decomposition, and a plurality of eigenvectors are obtained; the uniformity values of all the eigenvectors are calculated, and a target matrix is generated according to the two eigenvectors with the smallest uniformity values; the generated target matrix is subjected to clustering analysis, and a binaryzation gesture image is obtained according to a clustering analysis result; the maximum similarity value of the binaryzation gesture image and various prestored sample gesture images is obtained, and the channel is switched to a channel associated with the sample gesture image corresponding to the maximum similarity value. The invention further discloses a device for switching the channel through the gesture. Compared with the prior art, the user does not need to use a remote controller and can switch the channel only through the gesture, use of the user is facilitated, and the channel switching efficiency is improved.

Description

The method of gesture switching channels and device

Technical field

The present invention relates to graphics process field, particularly relate to method and the device of gesture switching channels.

Background technology

User uses a teleswitch and carries out remote control to TV, thus can realize the switching of channel.But the method for this switching channels that uses a teleswitch exists certain defect, such as, under the electricity of telepilot fault or telepilot such as to exhaust at the situation, user cannot use telepilot switching channels; Telepilot is away from user, and user first need take telepilot could switching channels.Therefore, the switching channels that uses a teleswitch, because being subject to the restriction of telepilot itself, causes the inconvenience that user uses, and the problem of inefficiency.

Foregoing, only for auxiliary understanding technical scheme of the present invention, does not represent and admits that foregoing is prior art.

Summary of the invention

Fundamental purpose of the present invention is that solution uses a teleswitch switching channels, causes the inconvenience that user uses, and the technical matters of inefficiency.

For achieving the above object, the invention provides a kind of method of gesture switching channels, the method for described gesture switching channels comprises the following steps:

Gather images of gestures, the described images of gestures gathered is converted to gray-scale map;

According to described gray-scale map, build Laplacian Matrix, feature decomposition is carried out to the described Laplacian Matrix built, obtains multiple proper vector;

According to the element in proper vector described in each, calculate the neat angle value of proper vector described in each, according to two proper vectors that described neat angle value is minimum, generate objective matrix;

Cluster analysis is carried out to the described objective matrix generated, according to the result of cluster analysis, the gray-scale value of the pixel in described images of gestures is set to 0 or 255, obtains the images of gestures of binaryzation;

The images of gestures obtaining described binaryzation and the maximum comparability value of each sample images of gestures prestored, switching channels is to the channel of sample images of gestures association corresponding to described maximum comparability value.

Preferably, described according to the element in proper vector described in each, calculate the neat angle value of proper vector described in each, according to two proper vectors that described neat angle value is minimum, the step generating objective matrix comprises:

Obtain the mean value of the value of element in proper vector described in each, calculate the summation of the value of element in described proper vector and the mean square deviation of described mean value, obtain described neat angle value;

Merge two proper vectors that described neat angle value is minimum, generate objective matrix.

Preferably, the step that the described described objective matrix to generating carries out cluster analysis comprises:

Steps A, selects arbitrarily two elements as the first cluster centre and the second cluster centre in described objective matrix;

Step B, calculate the distance value of each element and described first cluster centre in described objective matrix, described second cluster centre respectively, distance value with described first cluster centre is less than or equal to Elemental partition to the first cluster with the distance value of described second cluster centre, the distance value with described first cluster centre is greater than Elemental partition to the second cluster with the distance value of described second cluster centre;

Step C, calculates the mean value of all elements in described first cluster, regains the first cluster centre, calculates the mean value of all elements in described second cluster, regains the second cluster centre;

Continue to perform step B, until when described first cluster centre regained is identical with described second cluster centre with last described first cluster centre regained with described second cluster centre, export the result of cluster analysis; The result of described cluster analysis comprises: the element in the element in described first cluster that described first cluster centre finally regained is corresponding and described second cluster corresponding to described second cluster centre finally regained.

Preferably, the images of gestures of the described binaryzation of described acquisition comprises with the step of the maximum comparability value of each sample images of gestures prestored:

By the pixel of the images of gestures of described binaryzation and prestore each described in the pixel of sample images of gestures contrast one by one, obtain the quantity of the images of gestures of the described binaryzation pixel identical with sample images of gestures described in each, obtain the images of gestures of described binaryzation and the similarity of sample images of gestures described in each;

From the described similarity of the images of gestures of described binaryzation and sample images of gestures described in each, obtain maximum comparability value.

Preferably, described collection images of gestures, also comprises before the described images of gestures gathered is converted to the step of gray-scale map:

Under the environment of solid color background, gather multiple original sample images of gestures;

By gather each described in original sample images of gestures carry out binary conversion treatment, obtain described sample images of gestures;

Associate described sample images of gestures and channel, store described sample images of gestures.

In addition, for achieving the above object, the present invention also provides a kind of device of gesture switching channels, and the device of described gesture switching channels comprises:

Acquisition module, for gathering images of gestures, is converted to gray-scale map by the described images of gestures gathered;

Build module, for according to described gray-scale map, build Laplacian Matrix, feature decomposition is carried out to the described Laplacian Matrix built, obtains multiple proper vector;

Generation module, for according to the element in proper vector described in each, calculates the neat angle value of proper vector described in each, according to two proper vectors that described neat angle value is minimum, generates objective matrix;

Cluster Analysis module, for carrying out cluster analysis to the described objective matrix generated, according to the result of cluster analysis, being set to 0 or 255 by the gray-scale value of the pixel in described images of gestures, obtaining the images of gestures of binaryzation;

Handover module, for the maximum comparability value of the images of gestures obtaining described binaryzation and each sample images of gestures prestored, the channel that switching channels associates to the sample images of gestures that described maximum comparability value is corresponding.

Preferably, described generation module comprises:

Acquiring unit, for obtaining the mean value of the value of element in proper vector described in each, calculating the summation of the value of element in described proper vector and the mean square deviation of described mean value, obtaining described neat angle value;

Merge cells, for merging two minimum proper vectors of described neat angle value, generates objective matrix.

Preferably, described Cluster Analysis module comprises:

Initialization unit, for selecting arbitrarily two elements as the first cluster centre and the second cluster centre in described objective matrix;

First analytic unit, for calculating the distance value of each element and described first cluster centre in described objective matrix, described second cluster centre respectively, distance value with described first cluster centre is less than or equal to Elemental partition to the first cluster with the distance value of described second cluster centre, the distance value with described first cluster centre is greater than Elemental partition to the second cluster with the distance value of described second cluster centre;

Second analytic unit, for calculating the mean value of all elements in described first cluster, regains the first cluster centre, calculates the mean value of all elements in described second cluster, regains the second cluster centre;

First analytic unit, also for calculating the mean value of all elements in described first cluster at the second analytic unit, regain the first cluster centre, calculate the mean value of all elements in described second cluster, after regaining the second cluster centre, continue to perform and calculate each element and described first cluster centre in described objective matrix respectively, the distance value of described second cluster centre, distance value with described first cluster centre is less than or equal to Elemental partition to the first cluster with the distance value of described second cluster centre, distance value with described first cluster centre is greater than the step with Elemental partition to the second cluster of the distance value of described second cluster centre, until when described first cluster centre regained is identical with described second cluster centre with last described first cluster centre regained with described second cluster centre, export the result of cluster analysis, the result of described cluster analysis comprises: the element in the element in described first cluster that described first cluster centre finally regained is corresponding and described second cluster corresponding to described second cluster centre finally regained,

Binarization unit, for the result according to cluster analysis, is set to 0 or 255 by the gray-scale value of the pixel in described images of gestures, obtains the images of gestures of binaryzation.

Preferably, described handover module comprises:

First acquiring unit, for the images of gestures by described binaryzation pixel and prestore each described in the pixel of sample images of gestures contrast one by one, obtain the quantity of the images of gestures of the described binaryzation pixel identical with sample images of gestures described in each, obtain the images of gestures of described binaryzation and the similarity of sample images of gestures described in each;

Second acquisition unit, in the described similarity of the images of gestures from described binaryzation and sample images of gestures described in each, obtains maximum comparability value;

Switch unit, for the channel that switching channels associates to the sample images of gestures that described maximum comparability value is corresponding.

Preferably, the device of described gesture switching channels also comprises:

Acquisition module, under the environment of solid color background, gathers multiple original sample images of gestures;

Processing module, for by gather each described in original sample images of gestures carry out binary conversion treatment, obtain described sample images of gestures;

Relating module, for associating described sample images of gestures and channel, stores described sample images of gestures.

The present invention gathers images of gestures, and the described images of gestures gathered is converted to gray-scale map; According to described gray-scale map, build Laplacian Matrix, feature decomposition is carried out to the described Laplacian Matrix built, obtains multiple proper vector; According to the element in proper vector described in each, calculate the neat angle value of proper vector described in each, according to two proper vectors that described neat angle value is minimum, generate objective matrix; Cluster analysis is carried out to the described objective matrix generated, according to the result of cluster analysis, the gray-scale value of the pixel in described images of gestures is set to 0 or 255, obtains the images of gestures of binaryzation; The images of gestures obtaining described binaryzation respectively and the maximum comparability value of each sample images of gestures prestored, switching channels is to the channel of sample images of gestures association corresponding to described maximum comparability value.Compared to prior art, user of the present invention does not need to use a teleswitch, and only just need can be realized the switching of channel by gesture, convenient for users to use, and improves the efficiency of switching channels.

Accompanying drawing explanation

Fig. 1 is the schematic flow sheet of method first embodiment of gesture switching channels of the present invention;

Fig. 2 is the schematic flow sheet of method second embodiment of gesture switching channels of the present invention;

Fig. 3 is the regularity schematic diagram of proper vector;

Fig. 4 is the schematic flow sheet of method the 3rd embodiment of gesture switching channels of the present invention;

Fig. 5 is the schematic flow sheet of method the 4th embodiment of gesture switching channels of the present invention;

Fig. 6 is the schematic flow sheet of method the 5th embodiment of gesture switching channels of the present invention;

Fig. 7 is the high-level schematic functional block diagram of device first embodiment of gesture switching channels of the present invention;

Fig. 8 is the high-level schematic functional block diagram of device second embodiment of gesture switching channels of the present invention;

Fig. 9 is the high-level schematic functional block diagram of device the 3rd embodiment of gesture switching channels of the present invention;

Figure 10 is the high-level schematic functional block diagram of device the 4th embodiment of gesture switching channels of the present invention;

Figure 11 is the high-level schematic functional block diagram of device the 5th embodiment of gesture switching channels of the present invention.

The realization of the object of the invention, functional characteristics and advantage will in conjunction with the embodiments, are described further with reference to accompanying drawing.

Embodiment

Should be appreciated that specific embodiment described herein only in order to explain the present invention, be not intended to limit the present invention.

The invention provides a kind of method of gesture switching channels.

With reference to the schematic flow sheet that Fig. 1, Fig. 1 are method first embodiment of gesture switching channels of the present invention.

In method first embodiment of gesture switching channels of the present invention, the method comprises the following steps:

Step S10, gathers images of gestures, and the described images of gestures gathered is converted to gray-scale map;

Televisor gathers images of gestures by collecting device (as camera, camera etc.), and this collecting device can be arranged in a television set, or is placed on televisor, and the present invention is not construed as limiting.The images of gestures of collection is converted to gray-scale map by televisor, such as, by the images of gestures of collection from RGB (Red, Green, Blue, red, green, blue) color space convert to YUV (gray scale, the blue colourity to yellow color-separated, the red colourity to cyan color components) chrominance space, obtain gray-scale map.Certainly also have a lot of method images of gestures being converted to gray-scale map, do not repeat one by one at this.

It should be noted that the images of gestures of collection is not limited to singlehanded images of gestures, also can be the images of gestures of both hands.

Step S20, according to described gray-scale map, builds Laplacian Matrix, carries out feature decomposition, obtain multiple proper vector to the described Laplacian Matrix built;

This case can build Laplacian Matrix by following technological means, is only that an example helps to understand below, is not intended to limit the present invention.

The gray-scale map that step S10 obtains can represent with data set the following:

X＝{x ₁,x ₂,...,x _n}∈R ^d

Wherein, x _irepresent the arbitrfary point of data centralization, (1, n), n is data amount check to i ∈, and d represents data dimension, and R represents whole set of real numbers.

First with following formulae discovery scale parameter σ _i:

${\mathrm{\σ}}_i=\frac{1}{n}\underset{d=1}{\overset{n}{\mathrm{\Σ}}}\left|\right|x_i-x_d\left|\right|$

Wherein, x _darbitrfary point x in data level X _iapart from the d Neighbor Points of all the other each points, d=7, n is selected to be data amount check;

Again with following formulae discovery similarity matrix A:

A _ij＝exp(-||x _i-x _j|| ²/σ _iσ _j)，i,j∈(i,n)

Wherein, A _ijrepresent the arbitrary element of similarity matrix A, σ _i, σ _jrepresent data centralization arbitrfary point x respectively _iand x _jcorresponding scale parameter, || x _i-x _j|| represent some x _iand x _jeuclidean distance.

Finally build Laplacian Matrix with following formula:

L＝D ^-1/2AD ^1/2

Wherein D is diagonal matrix, the arbitrary element D on diagonal line _ijfor the elements A of i-th row of similarity matrix A _ijsummation, D _ijas the following formula:

$D_{\mathrm{ij}}=\underset{j=1}{\overset{n}{\mathrm{\Σ}}}{A}_{\mathrm{ij}}$

Carry out feature decomposition to the Laplacian Matrix built, obtain multiple proper vector, such as, the Laplacian Matrix feature decomposition of the capable n row of n is n eigenwert and n proper vector.Because the circular of feature decomposition is comparatively conventional, therefore not to repeat here.

Step S30, according to the element in proper vector described in each, calculates the neat angle value of proper vector described in each, according to two proper vectors that described neat angle value is minimum, generates objective matrix;

According to the value of the element in each proper vector, calculate the neat angle value of proper vector described in each, namely regularity refers to the degree of scatter of the value of element in proper vector.

Neat angle value is less, and show that in proper vector, the value of element is more close, the value of element is more close, then proper vector is more neat, then the information of images of gestures that comprises of this proper vector is also more, therefore, televisor rounds two minimum proper vectors of neat angle value, generates objective matrix.

Step S40, carries out cluster analysis to the described objective matrix generated, according to the result of cluster analysis, the gray-scale value of the pixel in described images of gestures is set to 0 or 255, obtains the images of gestures of binaryzation;

Televisor carries out cluster analysis to the objective matrix generated, and the method for cluster analysis comprises: K means clustering method, K central point clustering method, partition clustering method, hierarchy clustering method etc.

Televisor is by cluster analysis, and the element in objective matrix is divided into two classes, and this two class is prospect and the background of images of gestures.Wherein a class represents with 0, another kind ofly represents with 255, and then realizes the binaryzation of images of gestures.Now, in images of gestures, the TOTAL DIFFERENT COLOR of the color and background of prospect (hand), achieves the segmentation of images of gestures.

Step S50, the images of gestures obtaining described binaryzation and the maximum comparability value of each sample images of gestures prestored, switching channels is to the channel of sample images of gestures association corresponding to described maximum comparability value.

The images of gestures of televisor acquisition binaryzation and the maximum comparability value of each sample images of gestures prestored, the method obtaining similarity comprises: the method etc. of histogram matching, matrix decomposition, distinguished point based, and the method that pixel can also be used to contrast one by one calculates similarity.Obtain the sample images of gestures that maximum comparability value is corresponding, by channel switch interface, the channel that switching channels associates to sample images of gestures.

The present embodiment gathers images of gestures, and the described images of gestures gathered is converted to gray-scale map; According to the described images of gestures being converted to gray-scale map, build Laplacian Matrix, feature decomposition is carried out to the described Laplacian Matrix built, obtains multiple proper vector; According to the element in proper vector described in each, calculate the neat angle value of proper vector described in each, according to two proper vectors that described neat angle value is minimum, generate objective matrix; Carry out cluster analysis to the described objective matrix generated, according to the result of cluster analysis, the element in objective matrix is divided into two classes, this two class is prospect and the background of images of gestures.Wherein a class represents with 0, another kind ofly represents with 255, and then realizes the binaryzation of images of gestures; The images of gestures obtaining described binaryzation respectively and the maximum comparability value of each sample images of gestures prestored, switching channels is to the channel of sample images of gestures association corresponding to described maximum comparability value.Compared to prior art, the present embodiment user does not need to use a teleswitch, and only just need can be realized the switching of channel by gesture, convenient for users to use, and improves the efficiency of switching channels.On the other hand, the present embodiment also can make user when away from televisor, carries out the switching of channel.

With reference to the schematic flow sheet that Fig. 2, Fig. 2 are method second embodiment of gesture switching channels of the present invention.

In method second embodiment of gesture switching channels of the present invention, the difference of the present embodiment and the first embodiment is, the present embodiment is on the basis of the first embodiment, described according to the element in proper vector described in each, calculate the neat angle value of proper vector described in each, according to two proper vectors that described neat angle value is minimum, the step generating objective matrix comprises:

Step S31, obtains the mean value of the value of element in proper vector described in each, calculates the summation of the value of element in described proper vector and the mean square deviation of described mean value, obtains described neat angle value;

Step S32, merges two proper vectors that described neat angle value is minimum, generates objective matrix.

Televisor, according to the element in each proper vector, calculates the mean value A of the value of element in proper vector, utilizes the neat angle value of each proper vector of formulae discovery below:

$L=\underset{i=1}{\overset{i=n}{\mathrm{\Σ}}}{(E_i-A)}^2$

Wherein, L is the neat angle value of proper vector, E _ifor the value of element in proper vector, n is the number of element in proper vector, and i ∈ (1, n).

As shown in following formula, the formula of the neat angle value of above-mentioned calculating proper vector can also carry out suitable distortion as required, and the formula after distortion is as follows:

$L=\underset{i=1}{\overset{i=n}{\mathrm{\Σ}}}\sqrt{{(E_i-A)}^2}$

$L=\underset{i=1}{\overset{i=n}{\mathrm{\Σ}}}|E_i-A|$

Wherein, L is the neat angle value of proper vector, E _ifor the value of element in proper vector, n is the number of element in proper vector, and i ∈ (1, n).

After calculating the neat angle value of proper vector, obtain two proper vectors that neat angle value is minimum, merge this two proper vectors, generate objective matrix VR.Such as, two proper vectors are respectively A and B, merge two proper vectors, obtain objective matrix VR=[A, B].

In the present embodiment, after getting the element in each proper vector, can according to the regularity schematic diagram of each proper vector generating feature vector obtained, for two proper vectors intuitively seeing that according to this schematic diagram in proper vector, regularity is minimum.With reference to Fig. 3, wherein, a, b, c are respectively the regularity schematic diagram of three proper vectors, and horizontal ordinate X represents the numbering of a line or a row pixel in images of gestures, the value of the element in ordinate Y representation feature vector.The neat angle value being obtained a, b characteristic of correspondence vector by the formulae discovery of the neat angle value of above-mentioned calculating is minimum, also can see intuitively from Fig. 3, in a, b characteristic of correspondence vector, the value of element is compared to the value of element in c characteristic of correspondence vector, in a, b characteristic of correspondence vector, the value of element is more close, namely more neat, therefore, can see by the regularity schematic diagram of proper vector in Fig. 3 two proper vectors that in proper vector, neat angle value is minimum intuitively.

With reference to the schematic flow sheet that Fig. 4, Fig. 4 are method the 3rd embodiment of gesture switching channels of the present invention.

In method the 3rd embodiment of gesture switching channels of the present invention, the difference of the present embodiment and the first embodiment and the second embodiment is, the present embodiment is on the basis of the first embodiment and/or the second embodiment, and the step that the described described objective matrix to generating carries out cluster analysis comprises:

Step S41, selects arbitrarily two elements as the first cluster centre and the second cluster centre in described objective matrix;

Step S42, calculate the distance value of each element and described first cluster centre in described objective matrix, described second cluster centre respectively, distance value with described first cluster centre is less than or equal to Elemental partition to the first cluster with the distance value of described second cluster centre, the distance value with described first cluster centre is greater than Elemental partition to the second cluster with the distance value of described second cluster centre;

Step S43, calculates the mean value of all elements in described first cluster, regains the first cluster centre, calculates the mean value of all elements in described second cluster, regains the second cluster centre;

Step S44, continues to perform step S42, until when described first cluster centre regained is identical with described second cluster centre with last described first cluster centre regained with described second cluster centre, export the result of cluster analysis; The result of described cluster analysis comprises: the element in the element in described first cluster that described first cluster centre finally regained is corresponding and described second cluster corresponding to described second cluster centre finally regained.

Televisor selects two elements as the first cluster centre and the second cluster centre at random or arbitrarily in objective matrix; Utilize each element in formulae discovery objective matrix below respectively with the distance value of the first cluster centre and the second cluster centre:

$\mathrm{OD}=\sqrt{{(E-A)}^2}$

Wherein, OD is distance value, and E is the value of any one element in objective matrix, and A is the value of the first cluster centre or the value of the second cluster centre.

With reference to the schematic flow sheet that Fig. 5, Fig. 5 are method the 4th embodiment of gesture switching channels of the present invention.

In method the 4th embodiment of gesture switching channels of the present invention, the difference of the present embodiment and the first embodiment, the second embodiment, the 3rd embodiment is, the present embodiment is on the basis of the first embodiment, the second embodiment, the 3rd embodiment, and the images of gestures of the described binaryzation of described acquisition comprises with the step of the maximum comparability value of each sample images of gestures prestored:

Step S51, by the pixel of the images of gestures of described binaryzation and prestore each described in the pixel of sample images of gestures contrast one by one, obtain the quantity of the images of gestures of the described binaryzation pixel identical with sample images of gestures described in each, be the images of gestures of described binaryzation and the similarity of sample images of gestures described in each;

Step S52, from the described similarity of the images of gestures of described binaryzation and sample images of gestures described in each, obtains maximum comparability value.

The pixel of the images of gestures of binaryzation is numbered by televisor in order, is numbered in order by the pixel of each sample images of gestures equally.The images of gestures of binaryzation contrasted with the pixel of identical numbering in each sample images of gestures, identical pixel is labeled as " 1 ", and different pixels is labeled as " 0 ".The quantity of the images of gestures pixel identical with each sample images of gestures of statistics binaryzation, is the images of gestures of binaryzation and the similarity of each sample images of gestures.

From all similarities obtained, obtain maximum comparability value.

The present embodiment, by being contrasted one by one by the pixel of the pixel of the images of gestures of binaryzation with each sample images of gestures prestored, obtains the images of gestures of binaryzation and the similarity of each sample images of gestures.The method of the present embodiment can obtain maximum comparability value accurately.

With reference to the schematic flow sheet that Fig. 6, Fig. 6 are method the 5th embodiment of gesture switching channels of the present invention.

In method the 5th embodiment of gesture switching channels of the present invention, the difference of the present embodiment and the first embodiment, the second embodiment, the 3rd embodiment, the 4th embodiment is, the present embodiment is on the basis of the first embodiment, the second embodiment, the 3rd embodiment, the 4th embodiment, described collection images of gestures, comprises before the described images of gestures gathered is converted to the step of gray-scale map:

Step S60, under the environment of solid color background, gathers multiple original sample images of gestures;

Step S70, by gather each described in original sample images of gestures carry out binary conversion treatment, obtain described sample images of gestures;

Step S80, associates described sample images of gestures and channel, stores described sample images of gestures.

Televisor, under the environment of solid color background, gathers multiple original sample images of gestures, adopts solid color background to be to make follow-up image processing process simpler.Owing to gathering under the environment of solid color background, therefore, by original sample images of gestures from RGB (Red, Green, Blue, red, green, blue) color space convert is HSV (Hue, Saturation, Value, tone, saturation degree, brightness) chrominance space, binary-state threshold can be set, the gray-scale value that tone value in original sample images of gestures is more than or equal to the pixel of binary-state threshold is set to 255, the gray-scale value that tone value in original sample images of gestures is less than the pixel of binary-state threshold is set to 0, certainly, the gray-scale value that tone value in original sample images of gestures can also be more than or equal to the pixel of binary-state threshold is set to 0, the gray-scale value that tone value in original sample images of gestures is less than the pixel of binary-state threshold is set to 255.Now, the effect original sample images of gestures of black and white will be presented as sample images of gestures.

Sample images of gestures is associated with channel, generates relation one to one by the numbering of sample images of gestures and channel or name, frequency etc.Storing sample images of gestures is in order to follow-up use.

The present embodiment gathers original sample images of gestures under solid color background, simplifies the processing procedure to original sample images of gestures, improves efficiency.

The present invention further provides a kind of device of gesture switching channels.

With reference to the high-level schematic functional block diagram that Fig. 7, Fig. 7 are device first embodiment of gesture switching channels of the present invention.

In device first embodiment of gesture switching channels of the present invention, this device comprises:

Acquisition module 10, for gathering images of gestures, is converted to gray-scale map by the described images of gestures gathered;

Acquisition module 10 gathers images of gestures by collecting device (as camera, camera etc.), and this collecting device can be arranged in a television set, or is placed on televisor, and the present invention is not construed as limiting.The images of gestures of collection is converted to gray-scale map by acquisition module 10, such as, by the images of gestures of collection from RGB (Red, Green, Blue, red, green, blue) color space convert to YUV (gray scale, the blue colourity to yellow color-separated, the red colourity to cyan color components) chrominance space, obtain gray-scale map.Certainly also have a lot of method images of gestures being converted to gray-scale map, do not repeat one by one at this.

It should be noted that the images of gestures of collection is not limited to singlehanded images of gestures, also can be the images of gestures of both hands.

Build module 20, for according to described gray-scale map, build Laplacian Matrix, feature decomposition is carried out to the described Laplacian Matrix built, obtains multiple proper vector;

This case can build Laplacian Matrix by following technological means, is only that an example helps to understand below, is not intended to limit the present invention.

The gray-scale map that acquisition module 10 obtains can represent with data set the following:

X＝{x ₁,x ₂,...,x _n}∈R ^d

Wherein, x _irepresent the arbitrfary point of data centralization, (1, n), n is data amount check to i ∈, and d represents data dimension, and R represents whole set of real numbers.

First with following formulae discovery scale parameter σ _i:

${\mathrm{\σ}}_i=\frac{1}{n}\underset{d=1}{\overset{n}{\mathrm{\Σ}}}\left|\right|x_i-x_d\left|\right|$

Wherein, x _darbitrfary point x in data level X _iapart from the d Neighbor Points of all the other each points, d=7, n is selected to be data amount check;

Again with following formulae discovery similarity matrix A:

A _ij＝exp(-||x _i-x _j|| ²/σ _iσ _j)，i,j∈(i,n)

Wherein, A _ijrepresent the arbitrary element of similarity matrix A, σ _i, σ _jrepresent data centralization arbitrfary point x respectively _iand x _jcorresponding scale parameter, || x _i-x _j|| represent some x _iand x _jeuclidean distance.

Finally build Laplacian Matrix with following formula:

L＝D ^-1/2AD ^1/2

Wherein D is diagonal matrix, the arbitrary element Di on diagonal line _jfor the elements A of i-th row of similarity matrix A _ijsummation, as the following formula:

$D_{\mathrm{ij}}=\underset{j=1}{\overset{n}{\mathrm{\Σ}}}{A}_{\mathrm{ij}}$

Carry out feature decomposition to the Laplacian Matrix built, obtain multiple proper vector, such as, the Laplacian Matrix feature decomposition of the capable n row of n is n eigenwert and n proper vector.Because the circular of feature decomposition is comparatively conventional, therefore not to repeat here.

Generation module 30, for according to the element in proper vector described in each, calculates the neat angle value of proper vector described in each, according to two proper vectors that described neat angle value is minimum, generates objective matrix;

According to the value of the element in each proper vector, utilize the neat angle value of formulae discovery proper vector described in each, namely regularity refers to the degree of scatter of the value of element in proper vector.

Neat angle value is less, and show that in proper vector, the value of element is more close, the value of element is more close, then proper vector is more neat, then the information of images of gestures that comprises of this proper vector is also more, therefore, generation module 30 rounds two minimum proper vectors of neat angle value, generates objective matrix.

Cluster Analysis module 40, for carrying out cluster analysis to the described objective matrix generated, according to the result of cluster analysis, being set to 0 or 255 by the gray-scale value of the pixel in described images of gestures, obtaining the images of gestures of binaryzation;

Cluster Analysis module 40 is by cluster analysis, and the element in objective matrix is divided into two classes, and this two class is prospect and the background of images of gestures.Wherein a class represents with 0, another kind ofly represents with 255, and then realizes the binaryzation of images of gestures.Now, in images of gestures, the TOTAL DIFFERENT COLOR of the color and background of prospect (hand), achieves the segmentation of images of gestures.

Handover module 50, for the maximum comparability value of the images of gestures obtaining described binaryzation and each sample images of gestures prestored, the channel that switching channels associates to the sample images of gestures that described maximum comparability value is corresponding.

The images of gestures that handover module 50 obtains binaryzation and the maximum comparability value of each sample images of gestures prestored, the method obtaining similarity comprises: the method etc. of histogram matching, matrix decomposition, distinguished point based, and the method that pixel can also be used to contrast one by one calculates similarity.Obtain the sample images of gestures that maximum comparability value is corresponding, by channel switch interface, the channel that switching channels associates to sample images of gestures.

The present embodiment acquisition module 10 gathers images of gestures, and the described images of gestures gathered is converted to gray-scale map; Build module 20 according to the described images of gestures being converted to gray-scale map, build Laplacian Matrix, feature decomposition is carried out to the described Laplacian Matrix built, obtains multiple proper vector; Generation module 30, according to the element in proper vector described in each, calculates the neat angle value of proper vector described in each, according to two proper vectors that described neat angle value is minimum, generates objective matrix; Cluster Analysis module 40 carries out cluster analysis to the described objective matrix generated, and according to the result of cluster analysis, the element in objective matrix is divided into two classes, this two class is prospect and the background of images of gestures.Wherein a class represents with 0, another kind ofly represents with 255, and then realizes the binaryzation of images of gestures; The images of gestures that handover module 50 obtains described binaryzation respectively and the maximum comparability value of each sample images of gestures prestored, switching channels is to the channel of sample images of gestures association corresponding to described maximum comparability value.Compared to prior art, the present embodiment user does not need to use a teleswitch, and only just need can be realized the switching of channel by gesture, convenient for users to use, and improves the efficiency of switching channels.On the other hand, the present embodiment also can make user when away from televisor, carries out the switching of channel.

With reference to the high-level schematic functional block diagram that Fig. 8, Fig. 8 are device second embodiment of gesture switching channels of the present invention.

In device second embodiment of gesture switching channels of the present invention, the difference of the present embodiment and the first embodiment is, the present embodiment is on the basis of the first embodiment, and described generation module 30 comprises:

Acquiring unit 31, for obtaining the mean value of the value of element in proper vector described in each, calculating the summation of the value of element in described proper vector and the mean square deviation of described mean value, obtaining described neat angle value;

Merge cells 32, for merging two minimum proper vectors of described neat angle value, generates objective matrix.

Acquiring unit 31, according to the element in each proper vector, calculates the mean value A of the value of element in proper vector, utilizes the neat angle value of each proper vector of formulae discovery below:

$L=\underset{i=1}{\overset{i=n}{\mathrm{\Σ}}}{(E_i-A)}^2$

Wherein, L is the neat angle value of proper vector, E _ifor the value of element in proper vector, n is the number of element in proper vector, and i ∈ (1, n).

As shown in following formula, the formula of the neat angle value of above-mentioned calculating proper vector can also carry out suitable distortion as required, but all in protection scope of the present invention.

$L=\underset{i=1}{\overset{i=n}{\mathrm{\Σ}}}\sqrt{{(E_i-A)}^2}$

$L=\underset{i=1}{\overset{i=n}{\mathrm{\Σ}}}|E_i-A|$

Wherein, L is the neat angle value of proper vector, E _ifor the value of element in proper vector, n is the number of element in proper vector, and i ∈ (1, n).

After calculating the neat angle value of proper vector, merge cells 32 obtains two minimum proper vectors of neat angle value, merges this two proper vectors, generates objective matrix VR.Such as, two proper vectors are respectively A and B, merge two proper vectors, obtain objective matrix VR=[A, B].

With reference to the high-level schematic functional block diagram that Fig. 9, Fig. 9 are device the 3rd embodiment of gesture switching channels of the present invention.

In device the 3rd embodiment of gesture switching channels of the present invention, the difference of the present embodiment and the first embodiment and the second embodiment is, the present embodiment is on the basis of the first embodiment and/or the second embodiment, and described Cluster Analysis module 40 comprises:

Initialization unit 41, for selecting arbitrarily two elements as the first cluster centre and the second cluster centre in described objective matrix;

First analytic unit 42, for calculating the distance value of each element and described first cluster centre in described objective matrix, described second cluster centre respectively, distance value with described first cluster centre is less than or equal to Elemental partition to the first cluster with the distance value of described second cluster centre, the distance value with described first cluster centre is greater than Elemental partition to the second cluster with the distance value of described second cluster centre;

Second analytic unit 43, for calculating the mean value of all elements in described first cluster, regains the first cluster centre, calculates the mean value of all elements in described second cluster, regains the second cluster centre;

First analytic unit 42, also for calculating the mean value of all elements in described first cluster at the second analytic unit 43, regain the first cluster centre, calculate the mean value of all elements in described second cluster, after regaining the second cluster centre, continue to perform and calculate each element and described first cluster centre in described objective matrix respectively, the distance value of described second cluster centre, distance value with described first cluster centre is less than or equal to Elemental partition to the first cluster with the distance value of described second cluster centre, distance value with described first cluster centre is greater than the step with Elemental partition to the second cluster of the distance value of described second cluster centre, until when described first cluster centre regained is identical with described second cluster centre with last described first cluster centre regained with described second cluster centre, export the result of cluster analysis, the result of described cluster analysis comprises: the element in the element in described first cluster that described first cluster centre finally regained is corresponding and described second cluster corresponding to described second cluster centre finally regained,

Binarization unit 44, for the result according to cluster analysis, is set to 0 or 255 by the gray-scale value of the pixel in described images of gestures, obtains the images of gestures of binaryzation.

Initialization unit 41 selects two elements as the first cluster centre and the second cluster centre at random or arbitrarily in objective matrix; First analytic unit 42 utilize each element in formulae discovery objective matrix below respectively with the distance value of the first cluster centre and the second cluster centre:

$\mathrm{OD}=\sqrt{{(E-A)}^2}$

Wherein, OD is distance value, and E is the value of any one element in objective matrix, and A is the value of the first cluster centre or the value of the second cluster centre.

With reference to the handoff functionality module diagram that Figure 10, Figure 10 are device the 4th embodiment of gesture switching channels of the present invention.

In device the 4th embodiment of gesture switching channels of the present invention, the difference of the present embodiment and the first embodiment, the second embodiment, the 3rd embodiment is, the present embodiment is on the basis of the first embodiment, the second embodiment, the 3rd embodiment, and described handover module 50 comprises:

First acquiring unit 51, for the images of gestures by described binaryzation pixel and prestore each described in the pixel of sample images of gestures contrast one by one, obtain the quantity of the images of gestures of the described binaryzation pixel identical with sample images of gestures described in each, obtain the images of gestures of described binaryzation and the similarity of sample images of gestures described in each;

Second acquisition unit 52, in the described similarity of the images of gestures from described binaryzation and sample images of gestures described in each, obtains maximum comparability value;

Switch unit 53, for the channel that switching channels associates to the sample images of gestures that described maximum comparability value is corresponding.

The pixel of the images of gestures of binaryzation is numbered by the first acquiring unit 51 in order, is numbered in order by the pixel of each sample images of gestures equally.The images of gestures of binaryzation contrasted with the pixel of identical numbering in each sample images of gestures, identical pixel is labeled as " 1 ", and different pixels is labeled as " 0 ".The quantity of the images of gestures pixel identical with each sample images of gestures of statistics binaryzation, is the images of gestures of binaryzation and the similarity of each sample images of gestures.

Second acquisition unit 52, from all similarities obtained, obtains maximum comparability value.

The present embodiment, by being contrasted one by one by the pixel of the pixel of the images of gestures of binaryzation with each sample images of gestures prestored, obtains the images of gestures of binaryzation and the similarity of each sample images of gestures.

With reference to the high-level schematic functional block diagram that Figure 11, Figure 11 are device the 5th embodiment of gesture switching channels of the present invention.

In device the 5th embodiment of gesture switching channels of the present invention, the difference of the present embodiment and the first embodiment, the second embodiment, the 3rd embodiment, the 4th embodiment is, the present embodiment is on the basis of the first embodiment, the second embodiment, the 3rd embodiment, the 4th embodiment, and the device of described gesture switching channels also comprises:

Sample collection module 60, under the environment of solid color background, gathers multiple original sample images of gestures;

Processing module 70, for by gather each described in original sample images of gestures carry out binary conversion treatment, obtain described sample images of gestures;

Relating module 80, for associating described sample images of gestures and channel, stores described sample images of gestures.

Sample collection module 60, under the environment of solid color background, gathers multiple original sample images of gestures, adopts solid color background to be to make follow-up image processing process simpler.Owing to gathering under the environment of solid color background, therefore, processing module 70 by original sample images of gestures from RGB (Red, Green, Blue, red, green, blue) color space convert is HSV (Hue, Saturation, Value, tone, saturation degree, brightness) chrominance space, binary-state threshold can be set, the gray-scale value that tone value in original sample images of gestures is more than or equal to the pixel of binary-state threshold is set to 255, the gray-scale value that tone value in original sample images of gestures is less than the pixel of binary-state threshold is set to 0, certainly, the gray-scale value that tone value in original sample images of gestures can also be more than or equal to the pixel of binary-state threshold is set to 0, the gray-scale value that tone value in original sample images of gestures is less than the pixel of binary-state threshold is set to 255.Now, the effect original sample images of gestures of black and white will be presented as sample images of gestures.

Sample images of gestures associates with channel by relating module 80, generates relation one to one by the numbering of sample images of gestures and channel or name, frequency etc.Storing sample images of gestures is in order to follow-up use.

The present embodiment gathers original sample images of gestures under solid color background, simplifies the processing procedure to original sample images of gestures, improves efficiency.

These are only the preferred embodiments of the present invention; not thereby the scope of the claims of the present invention is limited; every utilize instructions of the present invention and accompanying drawing content to do equivalent structure or equivalent flow process conversion; or be directly or indirectly used in other relevant technical fields, be all in like manner included in scope of patent protection of the present invention.

Claims (10)

1. a method for gesture switching channels, is characterized in that, the method for described gesture switching channels comprises the following steps:

Gather images of gestures, the described images of gestures gathered is converted to gray-scale map;

According to described gray-scale map, build Laplacian Matrix, feature decomposition is carried out to the described Laplacian Matrix built, obtains multiple proper vector;

According to the element in proper vector described in each, calculate the neat angle value of proper vector described in each, according to two proper vectors that described neat angle value is minimum, generate objective matrix;

Cluster analysis is carried out to the described objective matrix generated, according to the result of cluster analysis, the gray-scale value of the pixel in described images of gestures is set to 0 or 255, obtains the images of gestures of binaryzation;

The images of gestures obtaining described binaryzation and the maximum comparability value of each sample images of gestures prestored, switching channels is to the channel of sample images of gestures association corresponding to described maximum comparability value.

2. the method for gesture switching channels as claimed in claim 1, it is characterized in that, described according to the element in proper vector described in each, calculate the neat angle value of proper vector described in each, according to two proper vectors that described neat angle value is minimum, the step generating objective matrix comprises:

Obtain the mean value of the value of element in proper vector described in each, calculate the summation of the value of element in described proper vector and the mean square deviation of described mean value, obtain described neat angle value;

Merge two proper vectors that described neat angle value is minimum, generate objective matrix.

3. the method for gesture switching channels as claimed in claim 1 or 2, is characterized in that, the step that the described described objective matrix to generating carries out cluster analysis comprises:

Steps A, selects arbitrarily two elements as the first cluster centre and the second cluster centre in described objective matrix;

Step B, calculate the distance value of each element and described first cluster centre in described objective matrix, described second cluster centre respectively, distance value with described first cluster centre is less than or equal to Elemental partition to the first cluster with the distance value of described second cluster centre, the distance value with described first cluster centre is greater than Elemental partition to the second cluster with the distance value of described second cluster centre;

Step C, calculates the mean value of all elements in described first cluster, regains the first cluster centre, calculates the mean value of all elements in described second cluster, regains the second cluster centre;

Continue to perform step B, until when described first cluster centre regained is identical with described second cluster centre with last described first cluster centre regained with described second cluster centre, export the result of cluster analysis; The result of described cluster analysis comprises: the element in the element in described first cluster that described first cluster centre finally regained is corresponding and described second cluster corresponding to described second cluster centre finally regained.

4. the method for gesture switching channels as claimed in claim 1 or 2, is characterized in that, the images of gestures of the described binaryzation of described acquisition comprises with the step of the maximum comparability value of each sample images of gestures prestored:

By the pixel of the images of gestures of described binaryzation and prestore each described in the pixel of sample images of gestures contrast one by one, obtain the quantity of the images of gestures of the described binaryzation pixel identical with sample images of gestures described in each, obtain the images of gestures of described binaryzation and the similarity of sample images of gestures described in each;

From the described similarity of the images of gestures of described binaryzation and sample images of gestures described in each, obtain maximum comparability value.

5. the method for gesture switching channels as claimed in claim 1 or 2, is characterized in that, described collection images of gestures, also comprises before the described images of gestures gathered is converted to the step of gray-scale map:

Under the environment of solid color background, gather multiple original sample images of gestures;

By gather each described in original sample images of gestures carry out binary conversion treatment, obtain described sample images of gestures;

Associate described sample images of gestures and channel, store described sample images of gestures.

6. a device for gesture switching channels, is characterized in that, the device of described gesture switching channels comprises:

Acquisition module, for gathering images of gestures, is converted to gray-scale map by the described images of gestures gathered;

Build module, for according to described gray-scale map, build Laplacian Matrix, feature decomposition is carried out to the described Laplacian Matrix built, obtains multiple proper vector;

Generation module, for according to the element in proper vector described in each, calculates the neat angle value of proper vector described in each, according to two proper vectors that described neat angle value is minimum, generates objective matrix;

Cluster Analysis module, for carrying out cluster analysis to the described objective matrix generated, according to the result of cluster analysis, being set to 0 or 255 by the gray-scale value of the pixel in described images of gestures, obtaining the images of gestures of binaryzation;

Handover module, for the maximum comparability value of the images of gestures obtaining described binaryzation and each sample images of gestures prestored, the channel that switching channels associates to the sample images of gestures that described maximum comparability value is corresponding.

7. the device of gesture switching channels as claimed in claim 6, it is characterized in that, described generation module comprises:

Acquiring unit, for obtaining the mean value of the value of element in proper vector described in each, calculating the summation of the value of element in described proper vector and the mean square deviation of described mean value, obtaining described neat angle value;

Merge cells, for merging two minimum proper vectors of described neat angle value, generates objective matrix.

8. the device of gesture switching channels as claimed in claims 6 or 7, it is characterized in that, described Cluster Analysis module comprises:

Initialization unit, for selecting arbitrarily two elements as the first cluster centre and the second cluster centre in described objective matrix;

First analytic unit, for calculating the distance value of each element and described first cluster centre in described objective matrix, described second cluster centre respectively, distance value with described first cluster centre is less than or equal to Elemental partition to the first cluster with the distance value of described second cluster centre, the distance value with described first cluster centre is greater than Elemental partition to the second cluster with the distance value of described second cluster centre;

Second analytic unit: for calculating the mean value of all elements in described first cluster, regain the first cluster centre, calculates the mean value of all elements in described second cluster, regains the second cluster centre;

First analytic unit, also for calculating the mean value of all elements in described first cluster at the second analytic unit, regain the first cluster centre, calculate the mean value of all elements in described second cluster, after regaining the second cluster centre, continue to perform and calculate each element and described first cluster centre in described objective matrix respectively, the distance value of described second cluster centre, distance value with described first cluster centre is less than or equal to Elemental partition to the first cluster with the distance value of described second cluster centre, distance value with described first cluster centre is greater than the step with Elemental partition to the second cluster of the distance value of described second cluster centre, until when described first cluster centre regained is identical with described second cluster centre with last described first cluster centre regained with described second cluster centre, export the result of cluster analysis, the result of described cluster analysis comprises: the element in the element in described first cluster that described first cluster centre finally regained is corresponding and described second cluster corresponding to described second cluster centre finally regained,

Binarization unit, for the result according to cluster analysis, is set to 0 or 255 by the gray-scale value of the pixel in described images of gestures, obtains the images of gestures of binaryzation.

9. the device of gesture switching channels as claimed in claims 6 or 7, it is characterized in that, described handover module comprises:

First acquiring unit, for the images of gestures by described binaryzation pixel and prestore each described in the pixel of sample images of gestures contrast one by one, obtain the quantity of the images of gestures of the described binaryzation pixel identical with sample images of gestures described in each, obtain the images of gestures of described binaryzation and the similarity of sample images of gestures described in each;

Second acquisition unit, in the described similarity of the images of gestures from described binaryzation and sample images of gestures described in each, obtains maximum comparability value;

Switch unit, for the channel that switching channels associates to the sample images of gestures that described maximum comparability value is corresponding.

10. the device of gesture switching channels as claimed in claims 6 or 7, it is characterized in that, the device of described gesture switching channels also comprises:

Sample collection module, under the environment of solid color background, gathers multiple original sample images of gestures;

Processing module, for by gather each described in original sample images of gestures carry out binary conversion treatment, obtain described sample images of gestures;

Relating module, for associating described sample images of gestures and channel, stores described sample images of gestures.