Revival monitoring method and system based on face detection

Abstract

The invention provides a revival monitoring method and system based on face detection, and the method comprises the steps: obtaining a gray image of the face of a patient, and carrying out the texture extraction and contour extraction of the gray image, so as to obtain a texture feature image and a contour feature image; fusing the texture feature image and the contour feature image into a facial feature image, and identifying a first reference eye feature image and a first reference mouth feature image; acquiring a real-time image, acquiring a real-time eye feature image and a real-time mouth feature image, and determining whether to perform awakening judgment by combining the first reference eye feature image and the first reference mouth feature image; if the awakening judgment is carried out, a plurality of frame images under continuous time frames are obtained, and then the eye change frequency and the mouth change frequency are determined; and judging whether the patient is awakened based on the eye change times and the mouth change times. Manual awakening monitoring is replaced by an automatic and intelligent mode, real-time monitoring on whether a patient is awakened or not is achieved, and meanwhile manpower is saved.

Description

Technical Field

[0001] The present invention relates to the technical field of image processing, and in particular to a method and system for awakening monitoring based on face detection. Background Art

[0002] General anesthesia, also known as general anesthesia, refers to the temporary suppression of the central nervous system caused by the inhalation of anesthetic drugs into the body through the respiratory tract, intravenous or intramuscular injection. The clinical manifestations are loss of consciousness, loss of systemic pain, amnesia, reflex suppression and skeletal muscle relaxation. The degree of suppression of the central nervous system is related to the drug concentration in the blood and can be controlled and regulated. This suppression is completely reversible. When the drug is metabolized or excreted from the body, the patient's consciousness and various reflexes gradually recover.

[0003] At the current stage, anesthesia resuscitation monitoring mainly relies on professional nursing staff to manually monitor the patient's physical condition, including various routine status monitoring of the patient, such as monitoring electrocardiogram, blood pressure, pulse, respiration, blood oxygen saturation, etc., to determine whether the patient is awake.

[0004] However, the traditional method of manually monitoring the patient's awakening after general anesthesia not only requires a lot of manpower to monitor the patient's awakening, but also fails to achieve the effect of real-time monitoring and is prone to monitoring lags. Summary of the invention

[0005] In view of the shortcomings of the prior art, the purpose of the present invention is to provide a method and system for awakening monitoring based on face detection, aiming to solve the technical problem in the prior art of manually monitoring whether a patient wakes up after general anesthesia, which is not only labor-intensive but also unable to achieve the effect of real-time monitoring.

[0006] In order to achieve the above objectives, in a first aspect, an embodiment of the present application provides a method for monitoring awakening based on face detection, comprising the following steps:

[0007] Acquire an initial image of the patient's face, perform grayscale processing on the initial image to obtain a grayscale image, perform texture extraction on the grayscale image to obtain a texture feature image, and perform contour extraction on the grayscale image to obtain a contour feature image;

[0008] Merging the texture feature image and the contour feature image into a facial feature image, and identifying a first reference eye feature image and a first reference mouth feature image in the facial feature image;

[0009] Acquire a real-time image of the patient's face, and acquire a real-time eye feature image and a real-time mouth feature image based on the real-time image, perform a similarity comparison between the real-time eye feature image and the first reference eye feature image, and perform a similarity comparison between the real-time mouth feature image and the first reference mouth feature image to determine whether to perform a reawakening judgment;

[0010] If the awakening judgment is performed, a plurality of frame images corresponding to the patient's face in continuous time frames are obtained, and a plurality of eye judgment images and a plurality of mouth judgment images in continuous time frames are obtained based on the frame images, and the number of eye changes and the number of mouth changes are obtained based on the plurality of eye judgment images and the plurality of mouth judgment images;

[0011] Whether the patient is awake is determined based on the number of eye changes and the number of mouth changes.

[0012] Compared with the prior art, the beneficial effects of the present invention are as follows: by acquiring the texture feature image from the initial image, the facial changes of the patient under different lighting conditions are taken into consideration, and the texture features with unchanged lighting are extracted; by acquiring the contour feature image from the initial image, the problem of incomplete capture of the detailed information of the facial contour by the texture feature image can be compensated, and then the two are merged into the facial feature image, ensuring accurate capture of the patient's face and improving the accuracy of awakening monitoring; by continuously monitoring the patient's eyes and mouth, the patient's awakening or not is automatically identified based on the changes in the eyes and / or mouth after awakening, and the manual awakening monitoring is replaced by an automated and intelligent method, so as to realize real-time monitoring of whether the patient is awake or not, and save manpower at the same time.

[0013] Furthermore, the step of performing texture extraction on the grayscale image to obtain a texture feature image includes:

[0014] Using a multi-scale and multi-directional filter to extract features from the grayscale image, so as to obtain real sub-features and imaginary sub-features in different directions at different scales;

[0015] Converting the real sub-features in different directions at the same scale into a first scale parameter, and converting the imaginary sub-features in different directions at the same scale into a second scale parameter;

[0016] weighted fusion of the first scale parameters at different scales into real features, and weighted fusion of the second scale parameters at different scales into imaginary features;

[0017] A texture feature image is acquired based on the real feature and the imaginary feature.

[0018] Furthermore, the formula for obtaining the texture feature image is:

[0019]

[0020] Among them, T w represents the texture feature image, T s represents the real part characteristic, T x Represents the imaginary part feature.

[0021] Furthermore, the step of performing contour extraction on the grayscale image to obtain a contour feature image includes:

[0022] Performing threshold segmentation on the grayscale image to obtain a binary image;

[0023] Perform edge detection on the binary image based on an edge detection operator to obtain a contour feature image.

[0024] Furthermore, the step of fusing the texture feature image and the contour feature image into a facial feature image comprises:

[0025] Respectively performing dimensionality reduction processing on the texture feature image and the contour feature image to obtain a first dimensionality reduction image and a second dimensionality reduction image;

[0026] A first weight and a second weight are set for the first reduced dimensionality image and the second reduced dimensionality image, respectively, and a facial feature image is generated based on the first weight, the first reduced dimensionality image, the second weight and the second reduced dimensionality image.

[0027] Furthermore, the step of performing a similarity comparison between the real-time eye feature image and the first reference eye feature image, and performing a similarity comparison between the real-time mouth feature image and the first reference mouth feature image to determine whether to perform a wake-up judgment includes:

[0028] Acquire a first similarity between the real-time eye feature image and the first reference eye feature image, and acquire a second similarity between the real-time mouth feature image and the first reference mouth feature image;

[0029] Comparing the first similarity with a first threshold, and comparing the second similarity with a second threshold;

[0030] If the first similarity is less than the first threshold and / or the second similarity is less than the second threshold, it is determined to perform a wake-up judgment.

[0031] Furthermore, the step of obtaining the number of eye changes and the number of mouth changes based on the plurality of eye judgment images and the plurality of mouth judgment images includes:

[0032] Comparing the similarity of the two eye judgment images in adjacent time frames to obtain a plurality of third similarity values;

[0033] Comparing a plurality of the third similarity values ​​with a third threshold value to determine the number of eye changes;

[0034] performing a similarity comparison on the two mouth judgment images in adjacent time frames to obtain a plurality of fourth similarity values;

[0035] A plurality of the fourth similarity values ​​are compared with a fourth threshold to determine the number of mouth changes.

[0036] Furthermore, the step of determining whether the patient is awake based on the number of eye changes and the number of mouth changes includes:

[0037] Comparing the number of eye changes with a blink threshold, and comparing the number of mouth changes with a mouth opening and closing threshold;

[0038] If the number of eye changes is greater than the blink threshold and / or the number of mouth changes is greater than the mouth opening and closing threshold, it is determined that the patient is awake.

[0039] In a second aspect, an embodiment of the present application provides a wake-up monitoring system based on facial detection, which is applied to the wake-up monitoring method based on facial detection as described in the first aspect above, and the system includes:

[0040] An extraction module, used for acquiring an initial image of the patient's face, performing grayscale processing on the initial image to acquire a grayscale image, performing texture extraction on the grayscale image to acquire a texture feature image, and performing contour extraction on the grayscale image to acquire a contour feature image;

[0041] A processing module, used for fusing the texture feature image and the contour feature image into a facial feature image, and identifying a first reference eye feature image and a first reference mouth feature image in the facial feature image;

[0042] a first judgment module, configured to obtain a real-time image of the patient's face, and obtain a real-time eye feature image and a real-time mouth feature image based on the real-time image, perform a similarity comparison between the real-time eye feature image and the first reference eye feature image, and perform a similarity comparison between the real-time mouth feature image and the first reference mouth feature image, so as to determine whether to perform a reawakening judgment;

[0043] an execution module, for obtaining a plurality of frame images corresponding to the patient's face in continuous time frames if the awakening judgment is performed, and obtaining a plurality of eye judgment images and a plurality of mouth judgment images in continuous time frames based on the frame images, and obtaining the number of eye changes and the number of mouth changes based on the plurality of eye judgment images and the plurality of mouth judgment images;

[0044] The second judgment module is used to judge whether the patient is awake based on the number of eye changes and the number of mouth changes.

[0045] In a third aspect, an embodiment of the present application provides a computer, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein when the processor executes the computer program, the wake-up monitoring method based on facial detection as described in the first aspect above is implemented.

[0046] In a fourth aspect, an embodiment of the present application provides a storage medium having a computer program stored thereon, which, when executed by a processor, implements the wake-up monitoring method based on facial detection as described in the first aspect above. BRIEF DESCRIPTION OF THE DRAWINGS

[0047] Figure 1 is a flow chart of a method for awakening monitoring based on face detection in a first embodiment of the present invention;

[0048] Figure 2 is a structural block diagram of a wake-up monitoring system based on face detection in a second embodiment of the present invention;

[0049] The following specific implementation manner will further illustrate the present invention in conjunction with the above-mentioned drawings. DETAILED DESCRIPTION

[0050] In order to facilitate the understanding of the present invention, the present invention will be described more fully below with reference to the relevant drawings. Several embodiments of the present invention are given in the drawings. However, the present invention can be implemented in many different forms and is not limited to the embodiments described herein. On the contrary, the purpose of providing these embodiments is to make the disclosure of the present invention more thorough and comprehensive.

[0051] It should be noted that when an element is referred to as being "fixed to" another element, it may be directly on the other element or there may be a central element. When an element is considered to be "connected to" another element, it may be directly connected to the other element or there may be a central element at the same time. The terms "vertical", "horizontal", "left", "right" and similar expressions used herein are for illustrative purposes only.

[0052] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as those commonly understood by those skilled in the art of the present invention. The terms used herein in the specification of the present invention are only for the purpose of describing specific embodiments and are not intended to limit the present invention. The term "and / or" used herein includes any and all combinations of one or more related listed items.

[0053] See also Figure 1 The first embodiment of the present invention provides a method for monitoring awakening based on face detection, comprising the following steps:

[0054] S10: acquiring an initial image of the patient's face, performing grayscale processing on the initial image to acquire a grayscale image, performing texture extraction on the grayscale image to acquire a texture feature image, and performing contour extraction on the grayscale image to acquire a contour feature image;

[0055] Specifically, the step S10 includes:

[0056] S110: extracting features from the grayscale image using a multi-scale and multi-directional filter to obtain real sub-features and imaginary sub-features in different directions at different scales;

[0057] It can be understood that the multi-scale multi-directional filter includes several scales, each scale includes several directions, and when processing the grayscale image, each pixel in the grayscale image has the real sub-features and the imaginary sub-features in different directions at several different scales.

[0058] S120: converting the real part sub-features in different directions at the same scale into a first scale parameter, and converting the imaginary part sub-features in different directions at the same scale into a second scale parameter;

[0059] Specifically, the real part sub-features in different directions at the same scale are compared with a coding threshold, the coding threshold is 0, the real part sub-features greater than 0 are coded as 1, and the real part sub-features less than or equal to 0 are coded as 0, thereby forming the first scale parameter. The second scale parameter is obtained in the same manner and will not be described in detail here.

[0060] S130: weighting and fusing the first scale parameters at different scales into real features, and weighting and fusing the second scale parameters at different scales into imaginary features;

[0061] S140: Acquire a texture feature image based on the real feature and the imaginary feature;

[0062] The formula for obtaining the texture feature image is:

[0063]

[0064] Among them, T w represents the texture feature image, T s represents the real part characteristic, T x Represents the imaginary part feature.

[0065] S150: performing threshold segmentation on the grayscale image to obtain a binary image;

[0066] Specifically, a first segmentation value is obtained based on the average grayscale of the grayscale image, and the grayscale image is segmented once using the first segmentation value to obtain a first segmented image and a second segmented image, and the average grayscales of the first segmented image and the second segmented image are respectively obtained. A second segmentation value is obtained based on the average of the average grayscales of the first segmented image and the second segmented image, and it is determined whether the second segmentation value is less than a set value. If the second segmentation value is less than the set value, the second segmentation value is determined as the final segmentation value, and the grayscale image is threshold segmented using the final segmentation value. If the second segmentation value is greater than the set value, the grayscale image is segmented several times in the same manner until the acquired segmentation value is less than the set value.

[0067] S160: performing edge detection on the binary image based on an edge detection operator to obtain a contour feature image;

[0068] Edge detection algorithms are now widely used and will not be described in detail here. In this embodiment, the edge detection is performed by using the LoG operator.

[0069] S20: fusing the texture feature image and the contour feature image into a facial feature image, and identifying a first reference eye feature image and a first reference mouth feature image in the facial feature image;

[0070] The step S20 comprises:

[0071] S210: performing dimensionality reduction processing on the texture feature image and the contour feature image respectively to obtain a first dimensionality reduction image and a second dimensionality reduction image;

[0072] In this embodiment, the dimensionality reduction processing is performed by PCA principal component analysis method, which maps high-dimensional features to low dimensions through linear changes. PCA principal component analysis method is a relatively conventional dimensionality reduction algorithm and will not be described in detail here. By performing the dimensionality reduction processing, the amount of data calculation can be effectively reduced, thereby improving the efficiency of acquiring the facial feature image.

[0073] S220: setting a first weight and a second weight for the first reduced dimension image and the second reduced dimension image, respectively, and generating a facial feature image based on the first weight, the first reduced dimension image, the second weight, and the second reduced dimension image;

[0074] The acquisition formula of the facial feature image is:

[0075] T m =α*T J1 +β*T J2 ,

[0076] Among them, T m represents the facial feature image, T J1 represents the first dimension reduction image, T J2 represents the second dimensionality reduction image, α represents the first weight, and β represents the second weight.

[0077] In this embodiment, a plurality of eye feature points and a plurality of mouth feature points are obtained from the facial feature image by using a facial feature point algorithm of a cascade regression tree model, and then the first reference eye feature image is formed by using the plurality of eye feature points, and the first reference mouth feature image is formed by using the plurality of mouth feature points. It can be understood that at this time, the patient's eyes and mouth are both in a closed state.

[0078] S30: acquiring a real-time image of the patient's face, and acquiring a real-time eye feature image and a real-time mouth feature image based on the real-time image, performing a similarity comparison between the real-time eye feature image and the first reference eye feature image, and performing a similarity comparison between the real-time mouth feature image and the first reference mouth feature image, so as to determine whether to perform a reawakening judgment;

[0079] The real-time eye feature image and the real-time mouth feature image are obtained in the same steps as the first reference eye feature image and the first reference mouth feature image, which will not be described in detail here.

[0080] Specifically, the step S30 includes:

[0081] S310: Acquire a first similarity between the real-time eye feature image and the first reference eye feature image, and acquire a second similarity between the real-time mouth feature image and the first reference mouth feature image;

[0082] S320: comparing the first similarity with a first threshold, and comparing the second similarity with a second threshold;

[0083] S330: If the first similarity is less than the first threshold and / or the second similarity is less than the second threshold, determine to perform a wake-up judgment;

[0084] It can be understood that if the first similarity is greater than the first threshold and the second similarity is greater than the second threshold, there is no change in the patient's eyes and mouth, and it is determined that no awakening judgment is performed, which further saves the amount of calculation and improves the efficiency of awakening monitoring.

[0085] S40: if the awakening judgment is performed, a plurality of frame images corresponding to the patient's face in continuous time frames are obtained, and a plurality of eye judgment images and a plurality of mouth judgment images in continuous time frames are obtained based on the frame images, and the number of eye changes and the number of mouth changes are obtained based on the plurality of eye judgment images and the plurality of mouth judgment images;

[0086] The steps of obtaining the eye judgment image and the mouth judgment image and the first reference eye feature map and the first reference mouth feature map are as follows: Figure 1 I will not elaborate on this here.

[0087] The step S40 comprises:

[0088] S410: performing a similarity comparison on the two eye judgment images in adjacent time frames to obtain a plurality of third similarity values;

[0089] S420: comparing a plurality of the third similarity values ​​with a third threshold value to determine the number of eye changes;

[0090] In the eye judgment image under every two adjacent time frames, the eye judgment image under the previous time frame is used as a reference to judge whether there is any change in the eye judgment image under the next time frame compared with the previous time frame. If there is a change, it is judged that a blinking action has been performed, and then the number of times the patient's eyes have blinked within a period of time is determined by determining the number of eye changes.

[0091] S430: performing a similarity comparison on the two mouth judgment images in adjacent time frames to obtain a plurality of fourth similarity values;

[0092] S440: Compare a plurality of the fourth similarity values ​​with a fourth threshold value to determine the number of mouth changes.

[0093] It can be understood that the number of mouth changes is the number of times the patient's mouth opens and closes.

[0094] S50: judging whether the patient is awake based on the number of eye changes and the number of mouth changes;

[0095] The step S50 comprises:

[0096] S510: comparing the number of eye changes with a blink threshold, and comparing the number of mouth changes with a mouth opening and closing threshold;

[0097] S520: If the number of eye changes is greater than the blink threshold and / or the number of mouth changes is greater than the mouth opening and closing threshold, it is determined that the patient is awake.

[0098] In this embodiment, the blink threshold is 5 times, and the mouth opening and closing threshold is 2 times.

[0099] By acquiring the texture feature image from the initial image, the facial changes of the patient under different lighting conditions are taken into consideration, and the texture features with unchanged lighting are extracted; by acquiring the contour feature image from the initial image, the problem of incomplete capture of detailed information of facial contours by the texture feature image can be compensated, and then the two are merged into the facial feature image, thereby ensuring accurate capture of the patient's face and improving the accuracy of awakening monitoring; by continuously monitoring the patient's eyes and mouth, the awakening of the patient is automatically identified based on the changes in the eyes and / or mouth after awakening, and the manual awakening monitoring is replaced by an automated and intelligent method, so as to realize real-time monitoring of whether the patient is awake or not, and save manpower at the same time.

[0100] See also Figure 2 The second embodiment of the present invention provides a wake-up monitoring system based on face detection, which is applied to the wake-up monitoring method based on face detection in the above embodiment, and will not be repeated here. As used below, the terms "module", "unit", "sub-unit", etc. can implement a combination of software and / or hardware for a predetermined function. Although the apparatus described in the following embodiments is preferably implemented in software, the implementation in hardware, or a combination of software and hardware, is also possible and conceivable.

[0101] The system comprises:

[0102] The extraction module 10 is used to obtain an initial image of the patient's face, perform grayscale processing on the initial image to obtain a grayscale image, perform texture extraction on the grayscale image to obtain a texture feature image, and perform contour extraction on the grayscale image to obtain a contour feature image;

[0103] The extraction module 10 comprises:

[0104] The first unit is used to extract features from the grayscale image using a multi-scale and multi-directional filter to obtain real sub-features and imaginary sub-features in different directions at different scales;

[0105] A second unit is used to convert the real sub-features in different directions at the same scale into a first scale parameter, and convert the imaginary sub-features in different directions at the same scale into a second scale parameter;

[0106] A third unit is used to weight and fuse the first scale parameters at different scales into real features, and weight and fuse the second scale parameters at different scales into imaginary features;

[0107] A fourth unit is used to obtain a texture feature image based on the real feature and the imaginary feature;

[0108] A fifth unit is used to perform threshold segmentation on the grayscale image to obtain a binary image;

[0109] A sixth unit is used to perform edge detection on the binary image based on an edge detection operator to obtain a contour feature image;

[0110] A processing module 20, configured to fuse the texture feature image and the contour feature image into a facial feature image, and identify a first reference eye feature image and a first reference mouth feature image in the facial feature image;

[0111] The processing module 20 includes:

[0112] A seventh unit is used to perform dimensionality reduction processing on the texture feature image and the contour feature image respectively to obtain a first dimensionality reduction image and a second dimensionality reduction image;

[0113] An eighth unit is used to set a first weight and a second weight for the first reduced dimension image and the second reduced dimension image, respectively, and generate a facial feature image based on the first weight, the first reduced dimension image, the second weight and the second reduced dimension image;

[0114] A first judgment module 30 is used to obtain a real-time image of the patient's face, and obtain a real-time eye feature image and a real-time mouth feature image based on the real-time image, perform a similarity comparison between the real-time eye feature image and the first reference eye feature image, and perform a similarity comparison between the real-time mouth feature image and the first reference mouth feature image to determine whether to perform a reawakening judgment;

[0115] The first determination module 30 includes:

[0116] A ninth unit, configured to obtain a first similarity between the real-time eye feature image and the first reference eye feature image, and to obtain a second similarity between the real-time mouth feature image and the first reference mouth feature image;

[0117] A tenth unit, configured to compare the first similarity with a first threshold, and compare the second similarity with a second threshold;

[0118] The eleventh unit is used to determine whether to perform a wake-up judgment if the first similarity is less than the first threshold and / or the second similarity is less than the second threshold.

[0119] The execution module 40 is used for obtaining a plurality of frame images corresponding to the patient's face in continuous time frames if the awakening judgment is performed, and obtaining a plurality of eye judgment images and a plurality of mouth judgment images in continuous time frames based on the frame images, and obtaining the number of eye changes and the number of mouth changes based on the plurality of eye judgment images and the plurality of mouth judgment images;

[0120] The execution module 40 includes:

[0121] A twelfth unit is used to compare the similarity of the two eye judgment images in adjacent time frames to obtain a plurality of third similarity values;

[0122] A thirteenth unit is used to compare a plurality of the third similarity values ​​with a third threshold value to determine the number of eye changes;

[0123] A fourteenth unit is used to compare the similarity of the two mouth judgment images in adjacent time frames to obtain a plurality of fourth similarity values;

[0124] A fifteenth unit, configured to compare a plurality of the fourth similarity values ​​with a fourth threshold value to determine the number of mouth changes;

[0125] A second judgment module 50, used for judging whether the patient is awake based on the number of eye changes and the number of mouth changes;

[0126] The second determination module 50 includes:

[0127] A sixteenth unit is used to compare the number of eye changes with a blink threshold, and to compare the number of mouth changes with a mouth opening and closing threshold;

[0128] The seventeenth unit is used to determine that the patient is awake if the number of eye changes is greater than the blink threshold and / or the number of mouth changes is greater than the mouth opening and closing threshold.

[0129] The present invention also provides a computer, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein when the processor executes the computer program, the wake-up monitoring method based on facial detection as described in the above technical solution is implemented.

[0130] The present invention also provides a storage medium on which a computer program is stored. When the computer program is executed by a processor, the awakening monitoring method based on face detection as described in the above technical solution is implemented.

[0131] In the description of this specification, the description with reference to the terms "one embodiment", "some embodiments", "examples", "specific examples", or "some examples" means that the specific features, structures, materials or characteristics described in conjunction with the embodiment or example are included in at least one embodiment or example of the present invention. In this specification, the schematic representation of the above terms does not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials or characteristics described may be combined in any one or more embodiments or examples in a suitable manner.

[0132] The above-mentioned embodiments only express several implementation methods of the present invention, and the description thereof is relatively specific and detailed, but it cannot be understood as limiting the scope of the patent of the present invention. It should be pointed out that, for ordinary technicians in this field, several variations and improvements can be made without departing from the concept of the present invention, which all belong to the protection scope of the present invention. Therefore, the protection scope of the patent of the present invention shall be subject to the attached claims.

Claims

1. A method for awakening monitoring based on face detection, characterized in that: The following steps are involved: Acquire an initial image of the patient's face, perform grayscale processing on the initial image to obtain a grayscale image, perform texture extraction on the grayscale image to obtain a texture feature image, and perform contour extraction on the grayscale image to obtain a contour feature image; Merging the texture feature image and the contour feature image into a facial feature image, and identifying a first reference eye feature image and a first reference mouth feature image in the facial feature image; Acquire a real-time image of the patient's face, and acquire a real-time eye feature image and a real-time mouth feature image based on the real-time image, perform a similarity comparison between the real-time eye feature image and the first reference eye feature image, and perform a similarity comparison between the real-time mouth feature image and the first reference mouth feature image to determine whether to perform a reawakening judgment; If the awakening judgment is performed, a plurality of frame images corresponding to the patient's face in continuous time frames are obtained, and a plurality of eye judgment images and a plurality of mouth judgment images in continuous time frames are obtained based on the frame images, and the number of eye changes and the number of mouth changes are obtained based on the plurality of eye judgment images and the plurality of mouth judgment images; Whether the patient is awake is determined based on the number of eye changes and the number of mouth changes.

2. The awakening monitoring method based on face detection according to claim 1, characterized in that: The step of extracting texture from the grayscale image to obtain a texture feature image comprises: Using a multi-scale and multi-directional filter to extract features from the grayscale image, so as to obtain real sub-features and imaginary sub-features in different directions at different scales; Converting the real sub-features in different directions at the same scale into a first scale parameter, and converting the imaginary sub-features in different directions at the same scale into a second scale parameter; weighted fusion of the first scale parameters at different scales into real features, and weighted fusion of the second scale parameters at different scales into imaginary features; A texture feature image is acquired based on the real feature and the imaginary feature.

3. The awakening monitoring method based on face detection according to claim 2, characterized in that: The formula for obtaining the texture feature image is: Among them, T w represents the texture feature image, T s represents the real part characteristic, T x Represents the imaginary part feature.

4. The awakening monitoring method based on face detection according to claim 1, characterized in that: The step of extracting the contour of the grayscale image to obtain the contour feature image comprises: Performing threshold segmentation on the grayscale image to obtain a binary image; Perform edge detection on the binary image based on an edge detection operator to obtain a contour feature image.

5. The awakening monitoring method based on face detection according to claim 1, characterized in that: The step of fusing the texture feature image and the contour feature image into a facial feature image comprises: Respectively performing dimensionality reduction processing on the texture feature image and the contour feature image to obtain a first dimensionality reduction image and a second dimensionality reduction image; A first weight and a second weight are set for the first reduced dimension image and the second reduced dimension image, respectively, and a facial feature image is generated based on the first weight, the first reduced dimension image, the second weight and the second reduced dimension image.

6. The awakening monitoring method based on face detection according to claim 1, characterized in that: The step of comparing the real-time eye feature image with the first reference eye feature image for similarity, and comparing the real-time mouth feature image with the first reference mouth feature image for similarity to determine whether to perform awakening judgment comprises: Acquire a first similarity between the real-time eye feature image and the first reference eye feature image, and acquire a second similarity between the real-time mouth feature image and the first reference mouth feature image; Comparing the first similarity with a first threshold, and comparing the second similarity with a second threshold; If the first similarity is less than the first threshold and / or the second similarity is less than the second threshold, it is determined to perform a wake-up judgment.

7. The method for awakening monitoring based on face detection according to claim 1, characterized in that: The step of obtaining the number of eye changes and the number of mouth changes based on the plurality of eye judgment images and the plurality of mouth judgment images comprises: Comparing the similarity of the two eye judgment images in adjacent time frames to obtain a plurality of third similarity values; Comparing a plurality of the third similarity values ​​with a third threshold value to determine the number of eye changes; performing a similarity comparison on the two mouth judgment images in adjacent time frames to obtain a plurality of fourth similarity values; A plurality of the fourth similarity values ​​are compared with a fourth threshold to determine the number of mouth changes.

8. The method for awakening monitoring based on face detection according to claim 1, characterized in that: The step of judging whether the patient is awake based on the number of eye changes and the number of mouth changes comprises: Comparing the number of eye changes with a blink threshold, and comparing the number of mouth changes with a mouth opening and closing threshold; If the number of eye changes is greater than the blink threshold and / or the number of mouth changes is greater than the mouth opening and closing threshold, it is determined that the patient is awake.

9. A face detection-based awakening monitoring system, applied to the face detection-based awakening monitoring method according to any one of claims 1 to 8, characterized in that: The system comprises: An extraction module, used for acquiring an initial image of the patient's face, performing grayscale processing on the initial image to acquire a grayscale image, performing texture extraction on the grayscale image to acquire a texture feature image, and performing contour extraction on the grayscale image to acquire a contour feature image; A processing module, used for fusing the texture feature image and the contour feature image into a facial feature image, and identifying a first reference eye feature image and a first reference mouth feature image in the facial feature image; a first judgment module, configured to obtain a real-time image of the patient's face, and obtain a real-time eye feature image and a real-time mouth feature image based on the real-time image, perform a similarity comparison between the real-time eye feature image and the first reference eye feature image, and perform a similarity comparison between the real-time mouth feature image and the first reference mouth feature image, so as to determine whether to perform a reawakening judgment; an execution module, for obtaining a plurality of frame images corresponding to the patient's face in continuous time frames if the awakening judgment is performed, and obtaining a plurality of eye judgment images and a plurality of mouth judgment images in continuous time frames based on the frame images, and obtaining the number of eye changes and the number of mouth changes based on the plurality of eye judgment images and the plurality of mouth judgment images; The second judgment module is used to judge whether the patient is awake based on the number of eye changes and the number of mouth changes.

10. A computer comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that: When the processor executes the computer program, the face detection-based wake-up monitoring method according to any one of claims 1 to 8 is implemented.

Images