Screen edge-based mobile phone playback living body attack identification method
A screen edge and attack recognition technology, applied in the field of image processing, can solve the problem that video playback attacks cannot be effectively avoided, and achieve the effect of strong generalization ability and high operating efficiency
- Summary
- Abstract
- Description
- Claims
- Application Information
AI Technical Summary
Problems solved by technology
Method used
Image
Examples
Embodiment 1
[0053] A mobile phone playback live attack recognition method based on the edge of the screen, this method is based on a recognition system, such as figure 1 As shown, the recognition system includes an input module, an image out-of-focus detection module, a screen edge detection module and a judgment output module;
[0054] Described input module carries out facial recognition to the photo that camera is captured, and after identifying human face, frame-selects human face image, extracts the human face image in the range of frame selection, and transmits human face image to image out-of-focus detection module;
[0055] The image out-of-focus detection module performs image out-of-focus detection on the face image input by the image transmission module, and transmits the face image that is determined not to be out of focus after the image out-of-focus detection to the screen edge detection module;
[0056] The screen edge detection module performs screen edge detection on the...
Embodiment 2
[0061] The present invention is on the basis of above-mentioned embodiment 1, in order to realize the present invention better, as figure 2 As shown, further, the image out-of-focus detection specifically includes the following steps:
[0062] Step S1: Gaussian blur denoising is performed on the face image;
[0063] Step S2: Grayscale the face image after Gaussian denoising;
[0064] Step S3: Filtering the gray-scaled face image through the Laplacian algorithm;
[0065] Step S4: Calculate the mean and variance of the filtered face image;
[0066] Step S5: Preset the variance threshold M for determining out-of-focus, compare the variance calculated in step S4 with the variance threshold M as a standard, and determine whether the face image input by the image transmission module is a face image that is not out of focus;
[0067] Step S6: Transmit the face image determined not to be out of focus to the screen edge detection module.
[0068] Working principle: Image out-of-fo...
Embodiment 3
[0071] The present invention is on the basis of above-mentioned embodiment 1-2, in order to realize the present invention better, as image 3 As shown, further, the screen edge detection specifically includes the following steps:
[0072] Step S7: performing canny edge extraction on the out-of-focus face image;
[0073] Step S8: performing Huffman straight line extraction on the face image after canny edge extraction;
[0074] Step S9: performing line segment filtering on the face image after Huffman line extraction;
[0075] Step S10: Count the number of line segments on the face image after line segment filtering;
[0076] Step S11: Transmit the statistical result of the line segment number counting in step S10 to the judgment output module.
[0077] Working principle: For the face image sent by the image out-of-focus detection module, after operations such as canny edge extraction, Huffman line extraction and line segment filtering, line segment filtering is to filter so...
PUM
Login to View More Abstract
Description
Claims
Application Information
Login to View More 


