A kind of television operating mode switching device and method based on human eye feature analysis
Technical field
The present invention relates to a kind of television operating mode switching device, the invention still further relates to a kind of television operating mode changing method based on human eye feature analysis based on human eye feature analysis.
Background technology
Television set is leisure, the learning tool that frequency of utilization is quite high in people's daily life, also is one of main power consumption household electrical appliance.People situation away from keyboard, sleeping when watching, that the child is not intended to unlatching occurs through regular meeting in the process of using television set, this is not only in the electricity wasting resource, also are in the useful life that consumes television set.
Existing television set power-saving technique can be implemented in the function that changes battery saving mode, sleeping shutdown under the unmanned state over to, but still has following defective: one, adopt equipment such as pyroscan, thermal infrared imager, cost is higher and judging nicety rate is low more; Two, can not discern the people does other things around television set state; Three, can not in battery saving mode, on-off state, efficiently switch.
Summary of the invention
At the problems referred to above, the invention provides a kind of television operating mode switching device based on human eye feature analysis, cost is lower and judging nicety rate is high; And can discern state that the people does other things around the television set, can in battery saving mode, on-off state, efficiently switch, energy-conserving and environment-protective, for this reason, the present invention also provides the television operating mode changing method.
A kind of television operating mode switching device based on human eye feature analysis, it is characterized in that: it comprises video camera, people's face detection module device, the human eye feature analysis modular device, background server and mode of operation switch controller, the output of described video camera connects described people's face detection module device, the output of described people's face detection module device connects described human eye feature analysis modular device and described background server respectively, the output of described human eye feature analysis modular device connects described background server, and the output of described background server connects described mode of operation switch controller.
A kind of television operating mode changing method based on human eye feature analysis, it is characterized in that: it is by human eye feature analysis, determines that beholder's human eye is opened entire, direction of visual lines, further determines the method that television operating mode switches.
It is further characterized in that:
Its concrete steps are as follows:
When (1) opening television set, the video camera unlatching is also carried out video acquisition;
Whether (2) people's face detection module device is analyzed the image information of obtaining from video camera, have people's face to occur to judge:
If unmanned face information sends to background server with analytical information;
If people's face information is arranged, detected people's face information is sent to the human eye feature analysis module;
(3) the human eye feature analysis modular device uses human eye location and human eye line-of-sight detection algorithm based on sight angle, and human eye information from video camera is analyzed, and determines that beholder's eyes are opened entire and direction of visual lines, and analysis result is sent to background server;
(4) background server carries out comprehensive statistics, judges the duration of above state and analyzes whether need to switch television operating mode;
The unmanned face information time that background server obtains surpasses preheating setting time, sends the standby request signal to the television operating mode switch controller;
The human eye closure state information time that background server obtains surpasses preheating setting time, sends the standby request signal to the television operating mode switch controller;
The information time of people's an eye line misalignment screen that background server obtains surpasses first preheating setting time, sends the battery saving mode request signal to the television operating mode switch controller;
The information time of people's an eye line misalignment screen that background server obtains surpasses second preheating setting time, sends the standby request signal to the television operating mode switch controller;
The background server statistics standby mode time surpasses preheating setting time, sends the shutdown request signal to the television operating mode switch controller;
(5) the television operating mode switch controller carries out the switching of the mode of operation of television set according to the mode of operation switching signal that gets access to.
Adopt said method of the present invention, can analyze the beholder by people's face detection module device and human eye detection module and whether sleep, sight line towards, switch television operating mode according to beholder's state, not only save resource, and improved television set user's quality of life.This method can also be applied on the various devices such as personal computer widely, has very high economic worth and social effect.
Description of drawings:
The structural representation of Fig. 1 for using among the present invention;
Fig. 2 is an operation principle block diagram of the present invention.
Embodiment:
The present invention includes video camera 1, people's face detection module device 5, human eye feature analysis modular device 2, background server 3 and mode of operation switch controller 4, the output of video camera 1 connects people's face detection module device 5, the output of people's face detection module device 5 connects human eye feature analysis modular device 2 and background server 3 respectively, the output of human eye feature analysis modular device 2 connects background server 3, and the output of background server 3 connects mode of operation switch controller 4.
A kind of television operating mode changing method based on human eye feature analysis is by human eye feature analysis, determines that beholder's human eye is opened entire, direction of visual lines, further determines the method that television operating mode switches.
Concrete job step is as follows:
(1) television set the time, video camera is opened and is also carried out video acquisition;
Whether (2) people's face detection module device 5, and the image information of obtaining from video camera 1 is analyzed, have people's face to occur to judge:
If unmanned face information sends to background server 3 with analytical information;
If people's face information is arranged, detected people's face information is sent to human eye feature analysis modular device 2;
(3) human eye feature analysis modular device 2 uses human eyes location and human eye line-of-sight detection algorithm based on sight angle, human eye information from video camera 1 is analyzed, determine that beholder's eyes open entire and direction of visual lines, and analysis result is sent to background server 3;
(4) background server 3 carries out comprehensive statistics, judges the duration of above state and analyzes whether need to switch television operating mode.
The unmanned face information time that background server 3 obtains surpasses preheating setting time, sends the standby request signal to television operating mode switch controller 4;
The human eye closure state information time that background server 3 obtains surpasses preheating setting time, sends the standby request signal to television operating mode switch controller 4;
The information time of people's an eye line misalignment screen that background server 3 obtains surpasses first preheating setting time, sends the battery saving mode request signal to television operating mode switch controller 4;
The information time of people's an eye line misalignment screen that background server 3 obtains surpasses second preheating setting time, sends the standby request signal to television operating mode switch controller 4;
The 3 statistics standby mode times of background server surpass preheating setting time, send the shutdown request signal to television operating mode switch controller 4;
(5) television operating mode switch controller 4 carries out the switching of the mode of operation of television set according to the mode of operation switching signal that gets access to.
In the said method, the image that 5 pairs of video cameras 1 of people's face detection module device collect among Fig. 2 carries out people's face and detects, based on the adaboost learning algorithm of knowing altogether, its step is to use off-line training and then design feature selecting grader efficiently in people's face detection module device, this grader selects to use the tree decision device, and whether judgement exists people's face.
Human eye feature analysis modular device 2 has used the eyes location algorithm in detecting among Fig. 2, based on the human eye line-of-sight detection algorithm of sight angle determine human eye the position, open and close the state, sight angle of closing.
At first use the human eye location algorithm based on human face structure, to detect position of human eye information, its step is as follows: be gray level image with the image processing that collects earlier, former gray level image done edge gray scale reinforcement; In conjunction with the position of human eye decision criteria of setting up according to people's face geometric properties, determine position of human eye then; Come the authenticity of the eyes that inspection goes out at last as symmetrical similarity with the bidimensional coefficient correlation, and utilize the gray integration projection of the binocular images vertical direction find, close the gray threshold of state to find the switching that to distinguish human eye, if human eye is open, further utilize integral projection to determine pupil position, the pupil is positioned at the peak value of integrogram; Algorithm steps is known all altogether on it.Utilize human eye line-of-sight detection algorithm afterwards based on sight angle, detect people's an eye line information, its step is as follows: after the upper eyelid circumference that utilizes the Hough transformation calculations to go out human eye reaches inside and outside angle point, iris center, detecting the method for having taked branch to handle on the direction of visual lines then: at first judge the visual angle, calculate the direction of visual lines of eyes again.Finally obtain beholder's direction of visual lines, algorithm steps is known all altogether on it.