A method and device for automatic image acquisition

A technology for automatic image acquisition, applied in character and pattern recognition, instruments, computer components, etc., can solve problems such as inconvenient and inaccurate operation, and achieve the effects of convenient operation, accurate shooting, and labor saving

Inactive Publication Date: 2008-07-09

VIMICRO CORP

1 Cites 96 Cited by

AI-Extracted Technical Summary

Problems solved by technology

[0004] The technical problem to be solved by the present invention is to provide a method and device for automatic image acquisitio...

Abstract

The invention provides an image automatic acquisition method, which comprises the following steps: (a) collecting an image; (b) analyzing the image to obtain image parameters; (c) judging whether the image meets pre-specified requirements on the image according to the image parameters, if yes, extracting the collected image according to scheduled rules, otherwise, prompting an adjustment and returning to step (a). The invention also provides an image automatic acquisition device, which comprises a main control unit for controlling and managing other units, an image collection unit for collecting the image, an image parameter extraction and judgement unit for analyzing the image to obtain the image parameters and judging whether the image meets the requirements, if no, a prompt is given, and the main control unit controls the image collection unit to collect the image again, if yes, the main control unit gives the prompt to an image generation unit to extract the image according to the scheduled rules.

Application Domain

Character and pattern recognition

Technology Topic

Image acquisitionControl unit +2

Image

Examples

Experimental program(1)

Example Embodiment

[0059] A feasible implementation for determining whether the resolution of the face region meets the requirements is as follows:

[0060] Set the minimum horizontal resolution of the face area to T fwidth , the minimum vertical resolution is T fheight , then if fwidth fwidth or height fheight , it is considered that the resolution of the face area does not meet the requirements, and prompts to reduce the distance between the subject and the image acquisition device (for example, ask the subject to be closer to the image acquisition device to shoot);

[0061] Step 230, according to the position and size of the face, determine whether the distance between the face and the edge of the image is appropriate, if not, give a prompt, go to step 110, otherwise, go to the next step;

[0062] A feasible implementation for determining whether the distance from the face to the edge of the image meets the requirements is as follows:

[0063] Define the ratio of the distance between the face and the left edge of the image to the width of the face as R l =fleft/fwidth, this parameter can be used to measure whether the distance between the face and the left edge of the image meets the requirements. If it is too close, the image will exceed the boundary range when cutting the image. set R l Threshold T Rl , judging if R l Rl Then it prompts that the face is too left, adjust it a little to the right; similarly, define the ratio of the distance between the face and the right edge of the image to the width of the face as R r =(W-fright)/width, this parameter can be used to measure whether the distance between the face and the right edge of the image meets the requirements. If it is too close, the image will exceed the boundary range when cutting the image. set R r Threshold T Rr , judging if R r Rr Then it prompts that the face is too right, adjust it a little to the left;

[0064] Step 240, judge whether the face area is too bright or too dark, if so, give a prompt, go to step 110, otherwise, go to the next step;

[0065] A feasible implementation of judging whether the face area is too bright or too dark is as follows:

[0066] The part of the face area whose horizontal coordinates are less than or equal to fx is defined as the left area of the face, and the part greater than or equal to fx is the right area of the face. Calculate the average brightness of all pixels in the left area of the face as avg l , the mean brightness of all pixels in the right area is avg r , define the difference between the left and right area brightness mean value as d avg =abs(avg l -avg r ), abs is the symbol for taking the absolute value. The average brightness of the face area is avg f =(avg l +avg r )/2. Set the maximum average brightness max of the face area avgf and the minimum average brightness min avgf , judge if avg fmax avgf The prompt is too bright, reduce the brightness of the face, if avg fmin avgf , the prompt is too dark, increase the brightness of the face.

[0067] Step 250, determine whether the face area is polarized, if so, give a prompt, go to step 110, otherwise, go to the next step;

[0068] A feasible implementation of determining whether the face area is polarized is as follows:

[0069] Set the brightness difference threshold between the left area and the right area of the face to T davg , if d avgT davg Then it indicates that the brightness difference between the left and right faces is too large, and the polarized light is serious, please correct it.

[0070] Step 260, using the eye positioning algorithm to determine the posture, if the posture does not meet the requirements (such as the eyes are not horizontal), give a prompt, go to step 110, otherwise, execute the next step;

[0071] A feasible embodiment for determining whether the eye posture meets the requirements is to perform eye positioning on the image on the basis of face detection (the eye positioning algorithm is also a mature algorithm, which can be realized by referring to the method mentioned in Chinese patent application 010500636. ) to find the position of the left eye and the right eye on the image. Assume that the left and right eye positions located in the image are respectively (x l , y l ), (x r , y r ), x l is the abscissa of the left eye, y l is the ordinate of the left eye, x r is the abscissa of the right eye, y r is the ordinate of the right eye, then the angle q between the eyes and the horizontal line BL for Among them, arctan(x) is the arc tangent function. Set the required eye inclination range [q′ min , q′ max ], then when q BLq′ max , it prompts that the right eye is too high, please keep both eyes level, when q BL

[0072] Step 270, according to the eye position image motion information between frames, determine whether to blink, if blink, give a prompt, go to step 110, otherwise, execute the next step;

[0073] A possible implementation of determining whether to blink is as follows:

[0074] Define the left eye area as a point (x l , y l ) is the center, and the width is ewidth=r ew *fwidth, the height is eheight=r eh an area of *fheight, where r ew , r eh To achieve the set ratio. Define the right eye area as a point (x r , y r ) is an area with a width of ewidth and a height of eheight. For two consecutively collected images, calculate the sum of the brightness differences of the corresponding pixels in the left eye area and the right eye area, assuming that the sum of the pixel gray levels of two adjacent frames in the left eye area is sum le and sum re. sum le and sum re Can be used to measure whether the left eye and right eye blink at this time, if sum le sume and sum re sume Then it is considered that the eyes have not blinked, and the photo can be taken, otherwise wait until sum le sume and sum re sume.

[0075] Step 280, perform facial expression recognition on the face, obtain the expression type, and determine whether the expression type meets the requirements according to the image type (for example, the photo type), if it does not meet the requirements, give a prompt, go to step 110, otherwise, go to step 150;

PUM

Description & Claims & Application Information

We can also present the details of the Description, Claims and Application information to help users get a comprehensive understanding of the technical details of the patent, such as background art, summary of invention, brief description of drawings, description of embodiments, and other original content. On the other hand, users can also determine the specific scope of protection of the technology through the list of claims; as well as understand the changes in the life cycle of the technology with the presentation of the patent timeline. Login to view more.