WO2021068486A1 - 基于图像识别的视力检测方法、装置、及计算机设备 - Google Patents

基于图像识别的视力检测方法、装置、及计算机设备 Download PDF

Info

Publication number
WO2021068486A1
WO2021068486A1 PCT/CN2020/087032 CN2020087032W WO2021068486A1 WO 2021068486 A1 WO2021068486 A1 WO 2021068486A1 CN 2020087032 W CN2020087032 W CN 2020087032W WO 2021068486 A1 WO2021068486 A1 WO 2021068486A1
Authority
WO
WIPO (PCT)
Prior art keywords
current
value
pixel
user
picture
Prior art date
Application number
PCT/CN2020/087032
Other languages
English (en)
French (fr)
Inventor
夏新
Original Assignee
深圳壹账通智能科技有限公司
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by 深圳壹账通智能科技有限公司 filed Critical 深圳壹账通智能科技有限公司
Publication of WO2021068486A1 publication Critical patent/WO2021068486A1/zh

Links

Images

Classifications

    • A—HUMAN NECESSITIES
    • A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
    • A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
    • A61B3/00—Apparatus for testing the eyes; Instruments for examining the eyes
    • A61B3/10—Objective types, i.e. instruments for examining the eyes independent of the patients' perceptions or reactions
    • A61B3/14—Arrangements specially adapted for eye photography
    • A—HUMAN NECESSITIES
    • A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
    • A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
    • A61B3/00—Apparatus for testing the eyes; Instruments for examining the eyes
    • A61B3/10—Objective types, i.e. instruments for examining the eyes independent of the patients' perceptions or reactions
    • A61B3/103—Objective types, i.e. instruments for examining the eyes independent of the patients' perceptions or reactions for determining refraction, e.g. refractometers, skiascopes
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T5/00—Image enhancement or restoration
    • G06T5/70—Denoising; Smoothing
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T7/00—Image analysis
    • G06T7/10—Segmentation; Edge detection
    • G06T7/13—Edge detection
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V40/00—Recognition of biometric, human-related or animal-related patterns in image or video data
    • G06V40/10—Human or animal bodies, e.g. vehicle occupants or pedestrians; Body parts, e.g. hands
    • G06V40/16—Human faces, e.g. facial parts, sketches or expressions
    • G06V40/161—Detection; Localisation; Normalisation
    • G06V40/166—Detection; Localisation; Normalisation using acquisition arrangements
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T2207/00—Indexing scheme for image analysis or image enhancement
    • G06T2207/10—Image acquisition modality
    • G06T2207/10004—Still image; Photographic image
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T2207/00—Indexing scheme for image analysis or image enhancement
    • G06T2207/30—Subject of image; Context of image processing
    • G06T2207/30196—Human being; Person
    • G06T2207/30201—Face

Definitions

  • This application relates to the field of artificial intelligence technology, and in particular to a vision detection method, device, computer equipment, and storage medium based on image recognition.
  • the human eye sees external things clearly based on the principle of convex lens imaging.
  • the image of the scene formed by the lens of the normal vision just falls on the retina, so that people can see the scene clearly.
  • the inventor realizes that at present, when detecting eyesight, the test is usually performed by viewing the eye chart by the examiner, and periodic detection cannot be performed through the intelligent terminal, resulting in low detection efficiency and limited detection data obtained.
  • the embodiments of the present application provide a vision detection method, device, computer equipment, and storage medium based on image recognition, aiming to solve the problem that when the vision is detected in the prior art, the tester usually uses the visual acuity chart to perform the test.
  • the intelligent terminal performs periodic detection, resulting in low detection efficiency and limited detection data obtained.
  • an embodiment of the present application provides a vision detection method based on image recognition, which includes:
  • a visual acuity curve is constructed correspondingly; wherein the visual acuity curve takes the time axis as the X axis, and The visual acuity value corresponding to each moment is the Y axis; the screening conditions include the screening time period and the visual acuity value; and
  • the vision curve is sent to the user terminal for display.
  • an embodiment of the present application provides a vision detection device based on image recognition, which includes:
  • the receiving unit is configured to receive the current user's face image uploaded by the user terminal if the difference between the current system time and the previous picture collection time is equal to the preset picture collection period;
  • the size obtaining unit is used to obtain the head length or head width in the current user's face image
  • the current spacing value obtaining unit is used to obtain the current spacing between the user and the screen according to the head length or the ratio of the head width to the corresponding side length of the standard camera rectangular frame and the standard spacing value of the standard camera rectangular frame value;
  • the current user vision value obtaining unit is used to call the pre-stored distance value and vision mapping table according to the current distance value to obtain the current user vision value corresponding to the current distance value;
  • the visual acuity curve construction unit is configured to construct a visual acuity curve corresponding to the current visual acuity value of the user and a set of historical visual acuity values satisfying preset screening conditions in the stored historical visual acuity value set of the user; wherein the visual acuity curve is based on time
  • the axis is the X axis, and the visual acuity value corresponding to each moment is the Y axis; the screening conditions include the screening time period and the visual acuity value;
  • the curve sending unit is used to send the vision curve to the user terminal for display.
  • an embodiment of the present application provides a computer device, which includes a memory, a processor, and a computer program stored on the memory and running on the processor, and the processor executes the computer
  • the program implements the vision detection method based on image recognition as described in the first aspect.
  • the embodiments of the present application also provide a computer-readable storage medium, wherein the computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the processor executes the above-mentioned first On the one hand, the vision detection method based on image recognition.
  • the embodiments of the present application provide a vision detection method, device, computer equipment, and storage medium based on image recognition, which realize that when using a smart terminal with a front camera, the user’s face images are automatically collected periodically to detect the user’s vision.
  • the visual acuity curve can be automatically generated, which not only improves the detection efficiency, but also obtains a large amount of detection data for data analysis.
  • FIG. 1 is a schematic diagram of an application scenario of a vision detection method based on image recognition provided by an embodiment of the application;
  • FIG. 2 is a schematic flowchart of a vision detection method based on image recognition provided by an embodiment of the application
  • FIG. 3 is a schematic diagram of a sub-flow of a vision detection method based on image recognition provided by an embodiment of the application;
  • FIG. 4 is a schematic diagram of another sub-flow of the vision detection method based on image recognition provided by an embodiment of the application;
  • FIG. 5 is a schematic diagram of another sub-flow of the vision detection method based on image recognition provided by an embodiment of the application;
  • FIG. 6 is a schematic block diagram of a vision detection device based on image recognition provided by an embodiment of the application.
  • FIG. 7 is a schematic block diagram of subunits of an image recognition-based vision detection device provided by an embodiment of the application.
  • FIG. 8 is a schematic block diagram of another subunit of the vision detection device based on image recognition provided by an embodiment of the application.
  • FIG. 9 is a schematic block diagram of another subunit of the vision detection device based on image recognition provided by an embodiment of the application.
  • FIG. 10 is a schematic block diagram of a computer device provided by an embodiment of the application.
  • FIG. 1 is a schematic diagram of an application scenario of a vision detection method based on image recognition provided by an embodiment of this application
  • FIG. 2 is a schematic flow chart of a vision detection method based on image recognition provided by an embodiment of this application.
  • the vision detection method based on image recognition is applied to the server, and the method is executed by application software installed in the server.
  • the method includes steps S110 to S160.
  • the front camera (that is, the camera that can easily capture the user's facial image) is set on the user terminal (the user terminal can also be understood as a smart terminal, such as a desktop computer, a tablet computer, a smart phone, etc.) Periodically collect the user's face image, and then the smart terminal uploads the collected user's face image to the server for image recognition to analyze the user's vision value.
  • the user terminal can also be understood as a smart terminal, such as a desktop computer, a tablet computer, a smart phone, etc.
  • the current system time is 9:00 am on January 1, 2018, and the time interval between 8:50 am on January 1, 2018 and the last picture collection time is 10 minutes, which is equal to the preset picture collection period of 10 minutes
  • the user terminal automatically uploads the current user's face image collected at the current system time to the server.
  • the server can estimate the distance between the human eye and the screen of the user's face image collected at each moment, and then convert it according to the distance between the human eye and the screen and the user's vision, so as to realize the cycle of the vision value Sexual monitoring.
  • the contour of the user's head may be obtained through edge detection to obtain the user's head length or head width.
  • step S120 includes:
  • S122 Perform Gaussian filtering on the grayscale picture to obtain a filtered picture
  • S124 Perform dual-threshold detection and edge connection on the processed picture to obtain a current face edge detection image corresponding to the current user face image;
  • the current user's face image uploaded by the user terminal is a color image in RGB format
  • the current user's face image may be Perform grayscale to obtain a grayscale picture.
  • Gaussian filtering is performed on the grayscale image, and the realization of image Gaussian filtering can be realized by two one-dimensional Gaussian kernels with two weightings respectively, that is, one-dimensional X-direction convolution is performed first, and then the obtained convolution result is performed again. Convolve in the Y direction, and finally get the filtered picture.
  • Gaussian filtering can effectively reduce the noise of grayscale images.
  • non-maximum value suppression processing can be performed.
  • non-maximum suppression is an edge sparse technology, and the effect of non-maximum suppression is "thin" edges.
  • the edge extracted only based on the gradient value is still very blurred.
  • Non-maximum suppression can help suppress all gradient values other than the local maximum to 0, so that edge detection can be performed more accurately.
  • double threshold detection and edge connection are performed at this time to obtain the current face edge detection image corresponding to the current user face image.
  • the double-threshold algorithm is used to detect the processed pictures, that is, two thresholds are calculated using the cumulative histogram. Anything greater than the high threshold must be an edge; anything less than the low threshold must not be an edge. If the detection result is greater than the low threshold but less than the high threshold, it is necessary to determine whether there is an edge pixel that exceeds the high threshold in the adjacent pixels of this pixel. If there is, the pixel is an edge, otherwise it is not an edge.
  • the above double threshold algorithm detects only the pixels on the edge in the processed picture.
  • edge pixels obtained after edge detection seldom completely describe an actual edge.
  • the edge pixels can be combined into meaningful edges using the connection method, so as to obtain the current face edge detection image corresponding to the current user face image.
  • the current face rectangular frame can just surround the current face edge detection image, and all four sides are the same as the current face.
  • Edge detection is the rectangular frame outside the image.
  • the length of the circumscribed rectangular frame corresponds to the width of the head, and the length of the corresponding head of the circumscribed rectangular frame.
  • step S121 includes:
  • R channel value R IJ , G channel value G IJ , B channel value B IJ and Gray IJ R IJ *0.299+G IJ corresponding to each pixel (I, J) in the current user's face image *0.587+B IJ *0.114, correspondingly calculate the gray value Gray IJ corresponding to each pixel (I, J); among them, Gray IJ represents the gray value of the pixel (I, J), and R IJ represents the pixel (I, J) corresponds to the R channel value, G IJ represents the G channel value corresponding to the pixel (I, J), and B IJ represents the B channel value corresponding to the pixel (I, J);
  • the server obtains the face image of the current user, it obtains the pixel matrix corresponding to the three channels R, G, and B of the face image of the current user, that is, each pixel in the face image of the current user.
  • Each of the three channels of R, G, and B corresponds to a pixel value.
  • Perform grayscale processing on the current user’s face image that is, combine the R channel value, G channel value, and B channel value corresponding to each pixel in the current user’s face image into a gray value to obtain the current
  • the current user's face image is read in as a gray-scale image to obtain a gray-scale matrix, where M and N are the length and width of the image.
  • M and N are the length and width of the image.
  • step S123 includes:
  • step S1234. Determine whether the current pixel is the last pixel in the filtered picture; if the current pixel is the last pixel in the filtered picture, perform step S1235; if the current pixel is not the filtered picture The last pixel in the middle, obtain the next pixel that is backward adjacent to the current pixel to update it as the current pixel, and return to step S1231;
  • the purpose of performing non-maximum value suppression for each pixel in the filtered image is to obtain the edge points of the image more accurately.
  • linear interpolation can be used between two adjacent pixels across the gradient direction to obtain the pixel gradient to be compared, that is, if the gradient intensity of the current pixel is greater than the two pixels along the positive and negative gradient directions of the current pixel
  • the current pixel value of the current pixel is retained; if the gradient intensity of the current pixel is less than the gradient intensity of the two pixels along the positive and negative gradient direction of the current pixel, the current pixel is Suppress, get suppressed pixels.
  • non-maximum value suppression it can help suppress all gradient values except the local maximum value to 0, so as to perform edge detection more accurately.
  • the camera of the smart terminal when the camera of the smart terminal collects the user’s facial image, it can virtually create a standard camera rectangular frame for face detection.
  • the standard camera rectangular frame is fixed and will not change between the user and the camera.
  • the size is changed by the change of the spacing.
  • the ratio of the head length corresponding to the current user’s face image or the head width to the standard length or standard width corresponding to the standard camera rectangle can be used to estimate the difference between the user and the screen. The distance between.
  • step S130 includes:
  • the ratio of the head length to the standard width and the standard distance value corresponding to the rectangular frame of the standard camera obtain the current distance value between the user and the screen; or according to the ratio of the head width to the standard length , And the standard distance value corresponding to the rectangular frame of the standard camera to obtain the current distance value between the user and the screen.
  • the distance between the user and the screen is obtained by the ratio of the head length corresponding to the current user's face image to the standard width corresponding to the standard camera rectangle, or the head corresponding to the current user's face image is obtained.
  • the ratio of the width of the part to the standard length corresponding to the rectangular frame of the standard camera obtains the distance between the user and the screen. In this way, the distance between the user and the screen can be effectively detected to estimate the vision value.
  • S140 According to the current distance value, and call a pre-stored distance value and vision mapping relationship table to obtain the current user's vision value corresponding to the current distance value.
  • the conversion can be performed according to a preset conversion curve between the distance value between the user and the screen and the vision value (the conversion curve can be understood as a mapping relationship between the distance value and the vision value), that is, the conversion between the distance value and the vision value
  • the conversion curve can be understood as a mapping relationship between the distance value and the vision value
  • Each distance value in the visual acuity mapping relationship table corresponds to a visual acuity value
  • the corresponding relationship between these distance values and the distance value and visual acuity value in the visual acuity mapping relationship table is valid data obtained through multiple experiments.
  • a distance between the user and the screen of 30cm corresponds to a vision of 5.0
  • a distance of 32cm between the user and the screen corresponds to a vision of 5.1, and so on. In this way, the distance between the user and the screen can be effectively converted into a visual acuity value for visual acuity monitoring.
  • the user's historical visual acuity value set stored in the server can be selected according to the filter condition, for example, the filter condition is that the filter time is from January 1, 2018 to January 5, 2018 and is not a null value.
  • the visual acuity value where the visual acuity value that is a null value is the embodiment of the user at the time point corresponding to the visual acuity curve when the user is not using the smart terminal at the current moment. That is, if the visual acuity value is a null value at a certain moment, it does not need to be drawn on the visual acuity curve.
  • the user’s visual acuity value change trend can be effectively monitored, so that whether an abnormal situation occurs can be known, and the user's use of the smart terminal can also be counted.
  • S160 Send the vision curve to the user terminal for display.
  • the visual acuity curve drawn according to the preset filtering conditions can be sent to the user terminal by the server for intuitive display on the user interaction interface on the user terminal , which can analyze the trend of vision changes.
  • step S160 the method further includes:
  • the visual acuity value at the previous moment and the visual acuity value at the next moment both have significant changes, for example, the visual acuity value at the current moment is 4.5, The visual acuity value at the previous moment is 5.0, and the visual acuity value at the next moment is 5.0, indicating that the visual acuity value at the current moment may be abnormal, and the user needs to be reminded to pay attention to vision protection.
  • the abnormal change interval of the visual acuity value can be highlighted (such as red) on the visual acuity curve, and this intuitive way can effectively prompt the customer.
  • This method realizes that when using a smart terminal with a front camera, it automatically periodically collects the user's face image to detect the user's vision, and can also automatically generate the vision curve, which not only improves the detection efficiency, but also obtains a large amount of detection data for data Analysis and use.
  • the embodiments of the present application also provide a vision detection device based on image recognition, which is used to execute any embodiment of the aforementioned vision detection method based on image recognition.
  • FIG. 6 is a schematic block diagram of a vision detection device based on image recognition provided by an embodiment of the present application.
  • the vision detection device 100 based on image recognition may be configured in a server.
  • the vision detection device 100 based on image recognition includes a receiving unit 110, a size acquiring unit 120, a current distance value acquiring unit 130, a current user vision value acquiring unit 140, a vision curve constructing unit 150, and a curve sending unit 160.
  • the receiving unit 110 is configured to receive the current user's face image uploaded by the user terminal if the difference between the current system time and the previous picture collection time is equal to the preset picture collection period.
  • the front camera (that is, the camera that can easily capture the user's facial image) is set on the user terminal (the user terminal can also be understood as a smart terminal, such as a desktop computer, a tablet computer, a smart phone, etc.) Periodically collect the user's face image, and then the smart terminal uploads the collected user's face image to the server for image recognition to analyze the user's vision value.
  • the user terminal can also be understood as a smart terminal, such as a desktop computer, a tablet computer, a smart phone, etc.
  • the current system time is 9:00 am on January 1, 2018, and the time interval between 8:50 am on January 1, 2018 and the last picture collection time is 10 minutes, which is equal to the preset picture collection period of 10 minutes
  • the user terminal automatically uploads the current user's face image collected at the current system time to the server.
  • the server can estimate the distance between the human eye and the screen of the user's face image collected at each moment, and then convert it according to the distance between the human eye and the screen and the user's vision, so as to realize the cycle of the vision value Sexual monitoring.
  • the size obtaining unit 120 is used to obtain the head length or the head width in the current user's face image.
  • the contour of the user's head may be obtained through edge detection to obtain the user's head length or head width.
  • the size obtaining unit 120 includes:
  • the gray-scale unit 121 is used to gray-scale the current user's face image to obtain a gray-scale picture
  • the filtering unit 122 is configured to perform Gaussian filtering on the grayscale picture to obtain a filtered picture
  • the non-extreme suppression unit 123 is configured to obtain the gradient value and direction of the filtered picture, and perform non-maximum suppression on the filtered picture to obtain a processed picture;
  • the edge detection unit 124 is configured to perform dual-threshold detection and connect edges on the processed picture to obtain a current face edge detection image corresponding to the current user face image;
  • the current face rectangle obtaining unit 125 is configured to obtain the corresponding current face rectangle frame according to the current face edge detection image, so as to obtain the head corresponding to the current user face image according to the current face rectangle frame.
  • the current user's face image uploaded by the user terminal is a color image in RGB format
  • the current user's face image may be Perform grayscale to obtain a grayscale picture.
  • Gaussian filtering is performed on the grayscale image, and the realization of image Gaussian filtering can be realized by two one-dimensional Gaussian kernels with two weightings respectively, that is, one-dimensional X-direction convolution is performed first, and then the obtained convolution result is performed again. Convolve in the Y direction, and finally get the filtered picture.
  • Gaussian filtering can effectively reduce the noise of grayscale images.
  • non-maximum value suppression processing can be performed.
  • non-maximum suppression is an edge sparse technology, and the effect of non-maximum suppression is "thin" edges.
  • the edge extracted only based on the gradient value is still very blurred.
  • Non-maximum suppression can help suppress all gradient values other than the local maximum to 0, so that edge detection can be performed more accurately.
  • double threshold detection and edge connection are performed at this time to obtain the current face edge detection image corresponding to the current user face image.
  • the double-threshold algorithm is used to detect the processed pictures, that is, two thresholds are calculated using the cumulative histogram. Anything greater than the high threshold must be an edge; anything less than the low threshold must not be an edge. If the detection result is greater than the low threshold but less than the high threshold, it is necessary to determine whether there is an edge pixel that exceeds the high threshold in the adjacent pixels of this pixel. If there is, the pixel is an edge, otherwise it is not an edge.
  • the above double threshold algorithm detects only the pixels on the edge in the processed picture.
  • edge pixels obtained after edge detection seldom completely describe an actual edge.
  • the edge pixels can be combined into meaningful edges using the connection method, so as to obtain the current face edge detection image corresponding to the current user face image.
  • the current face rectangular frame can just surround the current face edge detection image, and all four sides are the same as the current face.
  • Edge detection is the rectangular frame outside the image.
  • the length of the circumscribed rectangular frame corresponds to the width of the head, and the length of the corresponding head of the circumscribed rectangular frame.
  • the gray-scale unit 121 includes:
  • the RGB channel value acquisition unit 1211 is configured to acquire the R channel value, G channel value, and B channel value corresponding to each pixel (I, J) in the current user's face image; among them, the current user's face image slice
  • the total number of pixels is M*N, where M represents the total number of horizontal pixels in the current user's face image, N represents the total number of vertical pixels in the current user's face image, and the value of I
  • M represents the total number of horizontal pixels in the current user's face image
  • N represents the total number of vertical pixels in the current user's face image
  • the value of I The range is [0, M-1]
  • the value range of j is [0, N-1];
  • R IJ *0.299+G IJ *0.587+B IJ *0.114 correspondingly calculate the gray IJ corresponding to each pixel (I, J); among them, Gray IJ represents the gray of the pixel (I, J) Value
  • R IJ represents the R channel value corresponding to the pixel point (I, J)
  • G IJ represents the G channel value corresponding to the pixel point (I, J)
  • B IJ represents the B channel value corresponding to the pixel point (I, J);
  • the gray scale conversion unit 1213 is used to convert the R channel value, G channel value, and B channel value corresponding to each pixel (I, J) in the current user's face image into the corresponding gray value Gray ij to Get the corresponding grayscale picture.
  • the server obtains the face image of the current user, it obtains the pixel matrix corresponding to the three channels R, G, and B of the face image of the current user, that is, each pixel in the face image of the current user.
  • Each of the three channels of R, G, and B corresponds to a pixel value.
  • Perform grayscale processing on the current user’s face image that is, combine the R channel value, G channel value, and B channel value corresponding to each pixel in the current user’s face image into a gray value to obtain the current
  • the current user's face image is read in as a gray-scale image to obtain a gray-scale matrix, where M and N are the length and width of the image.
  • M and N are the length and width of the image.
  • the non-extreme value suppression unit 123 includes:
  • the gradient intensity comparison unit 1231 is used to compare the gradient intensity of the current pixel in the filtered picture with two pixels along the positive and negative gradient directions to determine whether the gradient intensity of the current pixel is greater than the current pixel
  • the gradient intensities of two pixels along the positive and negative gradient directions wherein, the initial value of the current pixel (i, j) is (0, 0), and the total number of pixels in the filtered picture is m*n, where m represents the total number of horizontal pixels in the filtered picture, n represents the total number of vertical pixels in the filtered picture, the value range of i is [1, m-1], j The value range is [1, n-1], and both m and n are natural numbers greater than 1;
  • the pixel retention unit 1232 is configured to retain the current pixel value of the current pixel if the gradient intensity of the current pixel is greater than the gradient intensities of the two pixels along the positive and negative gradient directions of the current pixel;
  • the pixel suppression unit 1233 is configured to suppress the current pixel to obtain the suppressed pixel if the gradient intensity of the current pixel is less than one of the gradient intensities of the two pixels along the positive and negative gradient directions of the current pixel;
  • the pixel end judging unit 1234 is used to judge whether the current pixel is the last pixel in the filtered picture; if the current pixel is the last pixel in the filtered picture, perform output of the current picture as the processed post The steps of the picture; if the current pixel is not the last pixel in the filtered picture, obtain the next pixel that is adjacent to the current pixel to update as the current pixel, and return to execute the filtered picture Compare the gradient intensity of the current pixel with the two pixels along the positive and negative gradient direction to determine whether the gradient intensity of the current pixel is greater than the gradient intensity of the two pixels along the positive and negative gradient direction of the current pixel A step of;
  • the current picture obtaining unit 1235 is configured to output the current picture as a processed picture.
  • the purpose of performing non-maximum value suppression for each pixel in the filtered image is to obtain the edge points of the image more accurately.
  • linear interpolation can be used between two adjacent pixels across the gradient direction to obtain the pixel gradient to be compared, that is, if the gradient intensity of the current pixel is greater than the two pixels along the positive and negative gradient directions of the current pixel
  • the current pixel value of the current pixel is retained; if the gradient intensity of the current pixel is less than the gradient intensity of the two pixels along the positive and negative gradient direction of the current pixel, the current pixel is Suppress, get suppressed pixels.
  • non-maximum value suppression it can help suppress all gradient values except the local maximum value to 0, so as to perform edge detection more accurately.
  • the current spacing value obtaining unit 130 is configured to obtain the current distance between the user and the screen according to the head length or the ratio of the head width to the corresponding side length of the standard camera rectangular frame and the standard spacing value of the standard camera rectangular frame. Spacing value.
  • the camera of the smart terminal when the camera of the smart terminal collects the user’s facial image, it can virtually create a standard camera rectangular frame for face detection.
  • the standard camera rectangular frame is fixed and will not change between the user and the camera.
  • the size is changed by the change of the spacing.
  • the ratio of the head length corresponding to the current user’s face image or the head width to the standard length or standard width corresponding to the standard camera rectangle can be used to estimate the difference between the user and the screen. The distance between.
  • the current distance value obtaining unit 130 includes:
  • the standard camera rectangular frame obtaining unit is configured to obtain a pre-stored standard camera rectangular frame, and the standard length or standard width corresponding to the standard camera rectangular frame;
  • the distance conversion unit is used to obtain the current distance value between the user and the screen according to the ratio of the head length to the standard width and the standard distance value corresponding to the rectangular frame of the standard camera; or according to the head width and the standard distance value
  • the ratio of the standard length and the standard spacing value corresponding to the rectangular frame of the standard camera are used to obtain the current spacing value between the user and the screen.
  • the distance between the user and the screen is obtained by the ratio of the head length corresponding to the current user's face image to the standard width corresponding to the standard camera rectangle, or the head corresponding to the current user's face image is obtained.
  • the ratio of the width of the part to the standard length corresponding to the rectangular frame of the standard camera obtains the distance between the user and the screen. In this way, the distance between the user and the screen can be effectively detected to estimate the vision value.
  • the current user vision value obtaining unit 140 is configured to call a pre-stored distance value and vision mapping table according to the current distance value to obtain the current user vision value corresponding to the current distance value.
  • the conversion can be performed according to a preset conversion curve between the distance value between the user and the screen and the vision value (the conversion curve can be understood as a mapping relationship between the distance value and the vision value), that is, the conversion between the distance value and the vision value
  • the conversion curve can be understood as a mapping relationship between the distance value and the vision value
  • Each distance value in the visual acuity mapping relationship table corresponds to a visual acuity value
  • the corresponding relationship between these distance values and the distance value and visual acuity value in the visual acuity mapping relationship table is valid data obtained through multiple experiments.
  • a distance between the user and the screen of 30cm corresponds to a vision of 5.0
  • a distance of 32cm between the user and the screen corresponds to a vision of 5.1, and so on. In this way, the distance between the user and the screen can be effectively converted into a visual acuity value for visual acuity monitoring.
  • the visual acuity curve construction unit 150 is configured to construct a visual acuity curve corresponding to the current visual acuity value of the user and the historical visual acuity value set satisfying the preset filtering condition in the stored historical visual acuity value set of the user; wherein the visual acuity curve is The time axis is the X axis, and the visual acuity value corresponding to each moment is the Y axis; the screening conditions include the screening time period and the value of the visual acuity value.
  • the user's historical visual acuity value set stored in the server can be selected according to the filter condition, for example, the filter condition is that the filter time is from January 1, 2018 to January 5, 2018 and is not a null value.
  • the visual acuity value where the visual acuity value that is a null value is the embodiment of the user at the time point corresponding to the visual acuity curve when the user is not using the smart terminal at the current moment. That is, if the visual acuity value is a null value at a certain moment, it does not need to be drawn on the visual acuity curve.
  • the user’s visual acuity value change trend can be effectively monitored, so that whether an abnormal situation occurs can be known, and the user's use of the smart terminal can also be counted.
  • the curve sending unit 160 is configured to send the vision curve to the user terminal for display.
  • the visual acuity curve drawn according to the preset filtering conditions can be sent to the user terminal by the server for intuitive display on the user interaction interface on the user terminal , which can analyze the trend of vision changes.
  • the vision detection device 100 based on image recognition further includes:
  • An abnormal interval obtaining unit configured to obtain an interval in the visual acuity curve where the visual acuity value change exceeds a preset visual acuity threshold in a unit time to form an abnormal visual acuity value change interval;
  • the highlight prompt unit is used to highlight the abnormal change interval of the vision value.
  • the visual acuity value at the previous moment and the visual acuity value at the next moment both have significant changes, for example, the visual acuity value at the current moment is 4.5, The visual acuity value at the previous moment is 5.0, and the visual acuity value at the next moment is 5.0, indicating that the visual acuity value at the current moment may be abnormal, and the user needs to be reminded to pay attention to vision protection.
  • the abnormal change interval of the visual acuity value can be highlighted (such as red) on the visual acuity curve, and this intuitive way can effectively prompt the customer.
  • the device realizes that when a smart terminal with a front camera is used, the user's face image is automatically collected periodically to detect the user's vision, and it can also automatically generate a vision curve, which not only improves the detection efficiency, but also obtains a large amount of detection data for data Analysis and use.
  • the above-mentioned vision detection apparatus based on image recognition can be implemented in the form of a computer program, and the computer program can be run on a computer device as shown in FIG. 10.
  • FIG. 10 is a schematic block diagram of a computer device according to an embodiment of the present application.
  • the computer device 500 is a server, and the server may be an independent server or a server cluster composed of multiple servers.
  • the computer device 500 includes a processor 502, a memory, and a network interface 505 connected through a system bus 501, where the memory may include a non-volatile storage medium 503 and an internal memory 504.
  • the non-volatile storage medium 503 can store an operating system 5031 and a computer program 5032.
  • the processor 502 can execute the vision detection method based on image recognition.
  • the processor 502 is used to provide computing and control capabilities, and support the operation of the entire computer device 500.
  • the internal memory 504 provides an environment for the operation of the computer program 5032 in the non-volatile storage medium 503.
  • the processor 502 can execute the vision detection method based on image recognition.
  • the network interface 505 is used for network communication, such as providing data information transmission.
  • the structure shown in FIG. 10 is only a block diagram of a part of the structure related to the solution of the present application, and does not constitute a limitation on the computer device 500 to which the solution of the present application is applied.
  • the specific computer device 500 may include more or fewer components than shown in the figure, or combine certain components, or have a different component arrangement.
  • the processor 502 is configured to run a computer program 5032 stored in a memory to implement the vision detection method based on image recognition disclosed in the embodiment of the present application.
  • the embodiment of the computer device shown in FIG. 10 does not constitute a limitation on the specific configuration of the computer device.
  • the computer device may include more or less components than those shown in the figure. Or some parts are combined, or different parts are arranged.
  • the computer device may only include a memory and a processor. In such an embodiment, the structures and functions of the memory and the processor are consistent with the embodiment shown in FIG. 10, and will not be repeated here.
  • the processor 502 may be a central processing unit (Central Processing Unit, CPU), and the processor 502 may also be other general-purpose processors, digital signal processors (Digital Signal Processors, DSPs), and special purpose processors.
  • Integrated circuit Application Specific Integrated Circuit, ASIC
  • ready-made programmable gate array Field-Programmable Gate Array, FPGA
  • the general-purpose processor may be a microprocessor or the processor may also be any conventional processor.
  • a computer-readable storage medium may be a non-volatile computer-readable storage medium, or may be a volatile computer-readable storage medium.
  • the computer readable storage medium stores a computer program, where the computer program is executed by a processor to implement the vision detection method based on image recognition disclosed in the embodiments of the present application.
  • the disclosed equipment, device, and method may be implemented in other ways.
  • the device embodiments described above are only illustrative.
  • the division of the units is only a logical function division. In actual implementation, there may be other division methods, or the units with the same function may be combined into one. Units, for example, multiple units or components can be combined or integrated into another system, or some features can be omitted or not implemented.
  • the displayed or discussed mutual coupling or direct coupling or communication connection may be indirect coupling or communication connection through some interfaces, devices or units, and may also be electrical, mechanical or other forms of connection.
  • the units described as separate components may or may not be physically separated, and the components displayed as units may or may not be physical units, that is, they may be located in one place, or they may be distributed on multiple network units. Some or all of the units may be selected according to actual needs to achieve the objectives of the solutions of the embodiments of the present application.
  • the functional units in the various embodiments of the present application may be integrated into one processing unit, or each unit may exist alone physically, or two or more units may be integrated into one unit.
  • the above-mentioned integrated unit can be implemented in the form of hardware or software functional unit.
  • the integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a storage medium.
  • the technical solution of this application is essentially or the part that contributes to the existing technology, or all or part of the technical solution can be embodied in the form of a software product, and the computer software product is stored in a storage medium. It includes several instructions to make a computer device (which may be a personal computer, a server, or a network device, etc.) execute all or part of the steps of the methods described in the various embodiments of the present application.
  • the aforementioned storage media include: U disk, mobile hard disk, read-only memory (ROM, Read-Only Memory), magnetic disk or optical disk and other media that can store program codes.

Landscapes

  • Engineering & Computer Science (AREA)
  • Health & Medical Sciences (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Physics & Mathematics (AREA)
  • Theoretical Computer Science (AREA)
  • General Health & Medical Sciences (AREA)
  • General Physics & Mathematics (AREA)
  • Ophthalmology & Optometry (AREA)
  • Heart & Thoracic Surgery (AREA)
  • Veterinary Medicine (AREA)
  • Public Health (AREA)
  • Biophysics (AREA)
  • Animal Behavior & Ethology (AREA)
  • Biomedical Technology (AREA)
  • Surgery (AREA)
  • Medical Informatics (AREA)
  • Molecular Biology (AREA)
  • Multimedia (AREA)
  • Human Computer Interaction (AREA)
  • Oral & Maxillofacial Surgery (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Image Processing (AREA)
  • Image Analysis (AREA)

Abstract

基于图像识别的视力检测方法、装置（100）、计算机设备（500）及存储介质。该方法包括获取当前用户人脸图像中的头部长度或是头部宽度（S120）；根据头部长度或是头部宽度与标准摄像头矩形框的对应边长之比，以获取用户与屏幕之间的当前间距值（S130）；根据当前间距值，及调用间距值与视力映射关系表，以获取对应的当前用户视力值（S140）；根据当前用户视力值及已存储的用户历史视力值集合中满足预设的筛选条件的历史视力值集合，对应构建视力曲线（S150）；将视力曲线发送至用户端以进行显示（S160）。该方法实现了使用带有前置摄像头的智能终端时自动周期性采集用户人脸图像以检测用户视力，还可自动生成视力曲线，不仅提高了检测效率，而且获取了海量的检测数据以供数据分析使用。

Description

基于图像识别的视力检测方法、装置、及计算机设备
本申请要求于2019年10月12日提交中国专利局、申请号为201910969805.X，发明名称为“基于图像识别的视力检测方法、装置、及计算机设备”的中国专利申请的优先权，其全部内容通过引用结合在本申请中。
技术领域
本申请涉及人工智能技术领域，尤其涉及一种基于图像识别的视力检测方法、装置、计算机设备及存储介质。
背景技术
人眼看清外界事物是基于凸透镜成像原理，正常视力眼球晶体所成景物图像恰好落在视网膜上，使得人能够清晰的可看到景物。
发明人意识到，目前，在检测视力时，通常采用检测者观看视力表的方式进行测试，无法通过智能终端进行周期性的检测，导致检测效率低，而且所获取的检测数据有限。
发明内容
本申请实施例提供了一种基于图像识别的视力检测方法、装置、计算机设备及存储介质，旨在解决现有技术中通检测视力时，通常采用检测者观看视力表的方式进行测试，无法通过智能终端进行周期性的检测，导致检测效率低，而且所获取的检测数据有限的问题。
第一方面，本申请实施例提供了一种基于图像识别的视力检测方法，其包括：
若当前系统时刻与上一图片采集时刻之差等于预设的图片采集周期，接收用户端所上传的当前用户人脸图像；
获取当前用户人脸图像中的头部长度或是头部宽度；
根据所述头部长度或是头部宽度与标准摄像头矩形框的对应边长之比以及标准摄像头矩形框的标准间距值，以获取用户与屏幕之间的当前间距值；
根据当前间距值，及调用预先存储的间距值与视力映射关系表，以获取与当前间距值对应的当前用户视力值；
根据所述当前用户视力值及已存储的用户历史视力值集合中满足预设的筛选条件的历史视力值集合，对应构建视力曲线；其中，所述视力曲线中以时间轴为X轴，且以各时刻对应的视力值为Y轴；所述筛选条件包括筛选时间段及视力值取值；以及
将所述视力曲线发送至用户端以进行显示。
第二方面，本申请实施例提供了一种基于图像识别的视力检测装置，其包括：
接收单元，用于若当前系统时刻与上一图片采集时刻之差等于预设的图片采集周期，接收用户端所上传的当前用户人脸图像；
尺寸获取单元，用于获取当前用户人脸图像中的头部长度或是头部宽度；
当前间距值获取单元，用于根据所述头部长度或是头部宽度与标准摄像头矩形框的对应边长之比以及标准摄像头矩形框的标准间距值，以获取用户与屏幕之间的当前间距值；
当前用户视力值获取单元，用于根据当前间距值，及调用预先存储的间距值与视力映射关系表，以获取与当前间距值对应的当前用户视力值；
视力曲线构建单元，用于根据所述当前用户视力值及已存储的用户历史视力值集合中满足预设的筛选条件的历史视力值集合，对应构建视力曲线；其中，所述视力曲线中以时间轴为X轴，且以各时刻对应的视力值为Y轴；所述筛选条件包括筛选时间段及视力值取值；以及
曲线发送单元，用于将所述视力曲线发送至用户端以进行显示。
第三方面，本申请实施例又提供了一种计算机设备，其包括存储器、处理器及存储在所述存储器上并可在所述处理器上运行的计算机程序，所述处理器执行所述计算机程序时实现上述第一方面所述的基于图像识别的视力检测方法。
第四方面，本申请实施例还提供了一种计算机可读存储介质，其中所述计算机可读存储介质存储有计算机程序，所述计算机程序当被处理器执行时使所述处理器执行上述第一方面所述的基于图像识别的视力检测方法。
本申请实施例提供了一种基于图像识别的视力检测方法、装置、计算机设备及存储介质，实现了使用带有前置摄像头的智能终端时自动周期性采集用户人脸图像以检测用户视力，还可自动生成视力曲线，不仅提高了检测效率，而且获取了海量的检测数据以供数据分析使用。
附图说明
图1为本申请实施例提供的基于图像识别的视力检测方法的应用场景示意图；
图2为本申请实施例提供的基于图像识别的视力检测方法的流程示意图；
图3为本申请实施例提供的基于图像识别的视力检测方法的子流程示意图；
图4为本申请实施例提供的基于图像识别的视力检测方法的另一子流程示意图；
图5为本申请实施例提供的基于图像识别的视力检测方法的另一子流程示意图；
图6为本申请实施例提供的基于图像识别的视力检测装置的示意性框图；
图7为本申请实施例提供的基于图像识别的视力检测装置的子单元示意性框图；
图8为本申请实施例提供的基于图像识别的视力检测装置的另一子单元示意性框图；
图9为本申请实施例提供的基于图像识别的视力检测装置的另一子单元示意性框图；
图10为本申请实施例提供的计算机设备的示意性框图。
具体实施方式
下面将结合本申请实施例中的附图，对本申请实施例中的技术方案进行清楚、完整地描述，显然，所描述的实施例是本申请一部分实施例，而不是全部的实施例。基于本申请中的实施例，本领域普通技术人员在没有做出创造性劳动前提下所获得的所有其他实施例，都属于本申请保护的范围。
请参阅图1和图2，图1为本申请实施例提供的基于图像识别的视力检测方法的应用场景示意图；图2为本申请实施例提供的基于图像识别的视力检测方法的流程示意图，该基于图像识别的视力检测方法应用于服务器中，该方法通过安装于服务器中的应用软件进行执行。
如图2所示，该方法包括步骤S110～S160。
S110、若当前系统时刻与上一图片采集时刻之差等于预设的图片采集周期，接收用户端所上传的当前用户人脸图像。
在本实施例中，是通过设置在用户端(用户端也可以理解为智能终端，如台式电脑、平板电脑、智能手机等)上的前置摄像头(即可以方便拍摄到用户面部图像的摄像头)周期性的采集用户人脸图像，之后由智能终端将所采集的用户人脸图像上传至服务器进行图像识别以分析用户视力值。
例如，当前系统时刻为2018年1月1日上午9:00，与上一图片采集时刻2018年1月1日上午8:50的时间间隔为10分钟，与预设的图片采集周期10分钟相等，此时由用户端自动将当前系统时刻所采集的当前用户人脸图像上传至服务器。服务器可以将采集得到的每一时刻的用户人脸图像进行人眼与屏幕之间的距离估算，之后根据人眼与屏幕之间的距离与使用者的视力进行换算，从而实现对视力值的周期性监测。
S120、获取当前用户人脸图像中的头部长度或是头部宽度。
在本实施例中，为了根据当前用户人脸图像获取头部长度或是头部宽度，可以通过边缘检测来获取用户的头部轮廓，以获取用户的头部长度或头部宽度。
在一实施例中，如图3所示，步骤S120包括：
S121、将所述当前用户人脸图像进行灰度化，得到灰度化图片；
S122、将所述灰度化图片进行高斯滤波，得到滤波后图片；
S123、获取所述滤波后图片的梯度值和方向，并将所述滤波后图片进行非极大值抑制，以得到处理后图片；
S124、将所述处理后图片进行双阈值检测及连接边缘，得到与所述当前的用户人脸图像对应的当前人脸边缘检测图像；
S125、根据所述当前人脸边缘检测图像获取对应的当前人脸矩形框，以根据所述当前人脸矩形框获取所述当前的用户人脸图像对应的头部长度或是头部宽度。
在本实施例中，由于用户端所上传的所述当前用户人脸图像是RGB格式的彩图，此时为了便于图像中人脸部分的轮廓，此时可以先将所述当前用户人脸图像进行灰度化，得到灰度化图片。
之后，对灰度化图片进行高斯滤波，其中图像高斯滤波的实现可以用两个一维高斯核分别两次加权实现，也就是先进行一维X方向卷积，得到的卷积结果再进行一维Y方向卷积，最终得到滤波后图片。当然也可以直接通过一个二维高斯核一次卷积实现。通过高斯滤波，能有效对灰度化图片进行降噪处理。
完成对所述当前用户人脸图像的高斯滤波后，可以进行非极大值抑制处理。其中，非极大值抑制是一种边缘稀疏技术，非极大值抑制的作用在于“瘦”边。对图像进行梯度计算后，仅仅基于梯度值提取的边缘仍然很模糊。而非极大值抑制则可以帮助将局部最大值之外的所有梯度值抑制为0，从而更准确的进行边缘检测。
对所述滤波后图片进行非极大值抑制得到处理后图片后，此时再进行双阈值检测及连接边缘，即可得到与所述当前的用户人脸图像对应的当前人脸边缘检测图像。其中，对处理后图片使用双阈值算法检测，即使用累计直方图计算两个阈值，凡是大于高阈值的一定是边缘；凡是小于低阈值的一定不是边缘。如果检测结果大于低阈值但又小于高阈值，则需判断这一个像素的邻接像素中有没有超过高阈值的边缘像素，如果有，则该像素就是边缘，否则就不是边缘。上述双阈值算法检测仅得到处理后图片中处在边缘上的像素点。噪声和不均匀的照明而产生的边缘间断的影响，使得经过边缘检测后得到的边缘像素点很少能完整地描绘实际的一条边缘。可以在使用边缘检测算法后，紧接着使用连接方法将边缘像素组合成有意义的边缘，从而得到与所述当前的用户人脸图像对应的当前人脸边缘检测图像。
最后，根据所述当前人脸边缘检测图像获取对应的当前人脸矩形框时，该当前人脸矩形框是恰好可将所述当前人脸边缘检测图像包围且四条边均与所述当前人脸边缘检测图像外切的矩形框。此时该外切的矩形框的长度即对应头部宽度，该外切的矩形框的对应头部长度。
在一实施例中，如图4所示，步骤S121包括：
S1211、获取所述当前用户人脸图像中每个像素点(I，J)对应的R通道值、G通道值、B通道值；其中，当前用户人脸图像片的总像素点个数为M*N个，其中M表示当前用户人脸图像片中横向像素点的总个数，N表示当前用户人脸图像片中纵向像素点的总个数，I的取值范围为[0，M-1]，j的取值范围为[0，N-1]；
S1212、根据所述当前用户人脸图像中每个像素点(I，J)对应的R通道值R IJ、G通道值G IJ、B通道值B IJ及Gray IJ＝R IJ*0.299+G IJ*0.587+B IJ*0.114，对应计算获取每个像素点(I，J)对应的灰度值Gray IJ；其中，Gray IJ表示像素点(I，J)的灰度值，R IJ表示像素点(I，J)对应的R通道值、G IJ表示像素点(I，J)对应的G通道值、B IJ表示像素点(I，J)对应的B通道值；
S1213、将所述当前用户人脸图像中每个像素点(I，J)对应的R通道值、G通道值、B通道值转化为对应的灰度值Gray ij，以得到对应的灰度化图片。
在本实施例中，服务器获取了所述当前用户人脸图像后，获取当前用户人脸图像在R、G、B三个通道分别对应的像素矩阵，即当前用户人脸图像中每个像素点在R、G、B三个通道分别对应一个像素值。对当前用户人脸图像进行灰度化处理，即是将当前用户人脸图像中每个 像素点对应的R通道值、G通道值、B通道值合并处理为一个灰度值，从而得到与当前用户人脸图像对应的灰度化图片，其中灰度化图片中每个像素点对应的灰度值Gray＝R*0.299+G*0.587+b*0.114。
即将当前用户人脸图像以灰度图像的形式读入，得到一个的灰度矩阵，其中M、N是图像的长、宽。这样读入比直接读入RGB彩色图像维度更低，同时没有明显损失图像信息。
在一实施例中，如图5所示，步骤S123包括：
S1231、将所述滤波后图片中当前像素点的梯度强度与沿正负梯度方向上的两个像素点进行比较，以判断当前像素点的梯度强度是否均大于当前像素点沿正负梯度方向上的两个像素点的梯度强度；其中，所述当前像素点(i，j)的初始值为(0,0)，所述滤波后图片中的总像素点个数为m*n个，其中m表示滤波后图片中横向像素点的总个数，n表示滤波后图片中纵向像素点的总个数，i的取值范围为[1，m-1]，j的取值范围为[1，n-1]，m、n均为大于1的自然数；
S1232、若当前像素点的梯度强度大于当前像素点沿正负梯度方向上的两个像素点的梯度强度，将当前像素点的当前像素值进行保留；
S1233、若当前像素点的梯度强度小于当前像素点沿正负梯度方向上的两个像素点的梯度强度其中一个梯度强度，将当前像素点进行抑制，得到抑制后像素点；
S1234、判断当前像素点是否为所述滤波后图片中最后一个像素点；若当前像素点为所述滤波后图片中最后一个像素点，执行步骤S1235；若当前像素点不为所述滤波后图片中最后一个像素点，获取与当前像素点向后相邻的下一像素点以更新作为当前像素点，返回执行步骤S1231；
S1235、输出当前图片作为处理后图片。
在本实施例中，即针对所述滤波后图片中各像素点进行非极大值抑制，是为了更加精确的获取图像的边缘点。具体的，在跨越梯度方向的两个相邻像素之间可使用线性插值来得到要比较的像素梯度，即若当前像素点的梯度强度大于当前像素点沿正负梯度方向上的两个像素点的梯度强度，将当前像素点的当前像素值进行保留；若当前像素点的梯度强度小于当前像素点沿正负梯度方向上的两个像素点的梯度强度其中一个梯度强度，将当前像素点进行抑制，得到抑制后像素点。通过非极大值抑制，则可以帮助将局部最大值之外的所有梯度值抑制为0，从而更准确的进行边缘检测。
S130、根据所述头部长度或是头部宽度与标准摄像头矩形框的对应边长之比以及标准摄像头矩形框的标准间距值，以获取用户与屏幕之间的当前间距值。
在本实施例中，由于智能终端的摄像头在采集用户的面部图像时，可以虚拟出面部检测的标准摄像头矩形框，该标准摄像头矩形框是固定不变的，不会随着用户与摄像头之间间距的变化而改变尺寸大小，通过所述当前用户人脸图像对应的头部长度或是头部宽度与该标准摄像头矩形框对应的标准长度或标准宽度的比例，即可估算出用户与屏幕之间的距离。
在一实施例中，步骤S130包括：
获取预先存储的标准摄像头矩形框，及所述标准摄像头矩形框对应的标准长度或标准宽度；
根据所述头部长度与所述标准宽度之比、及标准摄像头矩形框对应的标准间距值，获取用户与屏幕之间的当前间距值；或者根据所述头部宽度与所述标准长度之比、及标准摄像头矩形框对应的标准间距值，获取用户与屏幕之间的当前间距值。
即通过所述当前的用户人脸图像对应的头部长度与该标准摄像头矩形框对应的标准宽度之比获取用户与屏幕之间的距离，或是通过所述当前的用户人脸图像对应的头部宽度与该标准摄像头矩形框对应的标准长度度之比获取用户与屏幕之间的距离。通过这一方式，能有效检测出用户与屏幕之间的间距从而估算视力值。
S140、根据当前间距值，及调用预先存储的间距值与视力映射关系表，以获取与当前间距值对应的当前用户视力值。
在本实施例中，可根据预设的用户与屏幕之间的距离值与视力值的换算曲线(该换算曲线可以理解为间距值与视力映射关系表)来进行换算，也即在间距值与视力映射关系表中每一距离值都对应一个视力值，这些间距值与视力映射关系表中距离值与视力值的对应关系是通过多次试验得到的有效数据。例如用户与屏幕之间的距离为30cm对应5.0的视力，用户与屏幕之间的距离为32cm对应5.1的视力等。通过这一方式，可以有效的将用户与屏幕之间的距离值换算成视力值，从而进行视力监测。
S150、根据所述当前用户视力值及已存储的用户历史视力值集合中满足预设的筛选条件的历史视力值集合，对应构建视力曲线；其中，所述视力曲线中以时间轴为X轴，且以各时刻对应的视力值为Y轴；所述筛选条件包括筛选时间段及视力值取值；所述。
在本实施例中，可以在服务器中已存储的用户历史视力值集合中根据筛选条件，例如该筛选条件为筛选时间为2018年1月1日-2018年1月5日且不为空值的视力值，其中为空值的视力值是用户在当前时刻未使用智能终端时在视力曲线对应的时刻点上的体现。即若某一时刻视力值为空值，则不用绘制于该视力曲线图。通过绘制视力曲线，能有效监控用户的视力值变化趋势，从而可以获知是否有异常情况发生，也可以统计用户使用智能终端的时长。
S160、将所述视力曲线发送至用户端以进行显示。
在本实施例中，当在服务器中完成了根据预设的筛选条件而绘制的视力曲线后，可由服务器将所述视力曲线发送至用户端，以在用户端上的用户交互界面上进行直观展示，从而可以分析视力变化趋势。
在一实施例中，步骤S160之后还包括：
获取所述视力曲线中在单位时间内视力值变化超出预设的视力阈值的区间，以组成视力值异常变化区间；
将所述视力值异常变化区间进行高亮显示。
在本实施例中，若在所述视力曲线中检测到有某一时刻的视力值与前一时刻的视力值及后一时刻的视力值均发生显著变化，例如当前时刻的视力值为4.5，其前一时刻的视力值为5.0，其后一时刻的视力值为5.0，表示当前时刻的视力值可能存在异常，需要对用户进行提示以重视视力保护。此时，可将所述视力值异常变化区间在视力曲线上进行高亮显示(如红色)，通过这一直观的方式能对客户进行有效提示。
该方法实现了使用带有前置摄像头的智能终端时自动周期性采集用户人脸图像以检测用户视力，还可自动生成视力曲线，不仅提高了检测效率，而且获取了海量的检测数据以供数据分析使用。
本申请实施例还提供一种基于图像识别的视力检测装置，该基于图像识别的视力检测装置用于执行前述基于图像识别的视力检测方法的任一实施例。具体地，请参阅图6，图6是本申请实施例提供的基于图像识别的视力检测装置的示意性框图。该基于图像识别的视力检测装置100可以配置于服务器中。
如图6所示，基于图像识别的视力检测装置100包括接收单元110、尺寸获取单元120、当前间距值获取单元130、当前用户视力值获取单元140、视力曲线构建单元150、曲线发送单元160。
接收单元110，用于若当前系统时刻与上一图片采集时刻之差等于预设的图片采集周期，接收用户端所上传的当前用户人脸图像。
在本实施例中，是通过设置在用户端(用户端也可以理解为智能终端，如台式电脑、平板电脑、智能手机等)上的前置摄像头(即可以方便拍摄到用户面部图像的摄像头)周期性的采集用户人脸图像，之后由智能终端将所采集的用户人脸图像上传至服务器进行图像识别以分析用户视力值。
例如，当前系统时刻为2018年1月1日上午9:00，与上一图片采集时刻2018年1月1日上午8:50的时间间隔为10分钟，与预设的图片采集周期10分钟相等，此时由用户端自动将当前系统时刻所采集的当前用户人脸图像上传至服务器。服务器可以将采集得到的每一时 刻的用户人脸图像进行人眼与屏幕之间的距离估算，之后根据人眼与屏幕之间的距离与使用者的视力进行换算，从而实现对视力值的周期性监测。
尺寸获取单元120，用于获取当前用户人脸图像中的头部长度或是头部宽度。
在本实施例中，为了根据当前用户人脸图像获取头部长度或是头部宽度，可以通过边缘检测来获取用户的头部轮廓，以获取用户的头部长度或头部宽度。
在一实施例中，如图7所示，尺寸获取单元120包括：
灰度化单元121，用于将所述当前用户人脸图像进行灰度化，得到灰度化图片；
滤波单元122，用于将所述灰度化图片进行高斯滤波，得到滤波后图片；
非极值抑制单元123，用于获取所述滤波后图片的梯度值和方向，并将所述滤波后图片进行非极大值抑制，以得到处理后图片；
边缘检测单元124，用于将所述处理后图片进行双阈值检测及连接边缘，得到与所述当前的用户人脸图像对应的当前人脸边缘检测图像；
当前人脸矩形框获取单元125，用于根据所述当前人脸边缘检测图像获取对应的当前人脸矩形框，以根据所述当前人脸矩形框获取所述当前的用户人脸图像对应的头部长度或是头部宽度。
在本实施例中，由于用户端所上传的所述当前用户人脸图像是RGB格式的彩图，此时为了便于图像中人脸部分的轮廓，此时可以先将所述当前用户人脸图像进行灰度化，得到灰度化图片。
之后，对灰度化图片进行高斯滤波，其中图像高斯滤波的实现可以用两个一维高斯核分别两次加权实现，也就是先进行一维X方向卷积，得到的卷积结果再进行一维Y方向卷积，最终得到滤波后图片。当然也可以直接通过一个二维高斯核一次卷积实现。通过高斯滤波，能有效对灰度化图片进行降噪处理。
完成对所述当前用户人脸图像的高斯滤波后，可以进行非极大值抑制处理。其中，非极大值抑制是一种边缘稀疏技术，非极大值抑制的作用在于“瘦”边。对图像进行梯度计算后，仅仅基于梯度值提取的边缘仍然很模糊。而非极大值抑制则可以帮助将局部最大值之外的所有梯度值抑制为0，从而更准确的进行边缘检测。
对所述滤波后图片进行非极大值抑制得到处理后图片后，此时再进行双阈值检测及连接边缘，即可得到与所述当前的用户人脸图像对应的当前人脸边缘检测图像。其中，对处理后图片使用双阈值算法检测，即使用累计直方图计算两个阈值，凡是大于高阈值的一定是边缘；凡是小于低阈值的一定不是边缘。如果检测结果大于低阈值但又小于高阈值，则需判断这一个像素的邻接像素中有没有超过高阈值的边缘像素，如果有，则该像素就是边缘，否则就不是边缘。上述双阈值算法检测仅得到处理后图片中处在边缘上的像素点。噪声和不均匀的照明而产生的边缘间断的影响，使得经过边缘检测后得到的边缘像素点很少能完整地描绘实际的一条边缘。可以在使用边缘检测算法后，紧接着使用连接方法将边缘像素组合成有意义的边缘，从而得到与所述当前的用户人脸图像对应的当前人脸边缘检测图像。
最后，根据所述当前人脸边缘检测图像获取对应的当前人脸矩形框时，该当前人脸矩形框是恰好可将所述当前人脸边缘检测图像包围且四条边均与所述当前人脸边缘检测图像外切的矩形框。此时该外切的矩形框的长度即对应头部宽度，该外切的矩形框的对应头部长度。
在一实施例中，如图8所示，灰度化单元121包括：
RGB通道值获取单元1211，用于获取所述当前用户人脸图像中每个像素点(I，J)对应的R通道值、G通道值、B通道值；其中，当前用户人脸图像片的总像素点个数为M*N个，其中M表示当前用户人脸图像片中横向像素点的总个数，N表示当前用户人脸图像片中纵向像素点的总个数，I的取值范围为[0，M-1]，j的取值范围为[0，N-1]；
灰度值计算单元1212，用于根据所述当前用户人脸图像中每个像素点(I，J)对应的R通道值R IJ、G通道值G IJ、B通道值B IJ及Gray IJ＝R IJ*0.299+G IJ*0.587+B IJ*0.114，对应计算获取每个像素点(I，J)对应的灰度值Gray IJ；其中，Gray IJ表示像素点(I，J)的灰度值，R IJ 表示像素点(I，J)对应的R通道值、G IJ表示像素点(I，J)对应的G通道值、B IJ表示像素点(I，J)对应的B通道值；
灰度转化单元1213，用于将所述当前用户人脸图像中每个像素点(I，J)对应的R通道值、G通道值、B通道值转化为对应的灰度值Gray ij，以得到对应的灰度化图片。
在本实施例中，服务器获取了所述当前用户人脸图像后，获取当前用户人脸图像在R、G、B三个通道分别对应的像素矩阵，即当前用户人脸图像中每个像素点在R、G、B三个通道分别对应一个像素值。对当前用户人脸图像进行灰度化处理，即是将当前用户人脸图像中每个像素点对应的R通道值、G通道值、B通道值合并处理为一个灰度值，从而得到与当前用户人脸图像对应的灰度化图片，其中灰度化图片中每个像素点对应的灰度值Gray＝R*0.299+G*0.587+b*0.114。
即将当前用户人脸图像以灰度图像的形式读入，得到一个的灰度矩阵，其中M、N是图像的长、宽。这样读入比直接读入RGB彩色图像维度更低，同时没有明显损失图像信息。
在一实施例中，如图9所示，非极值抑制单元123包括：
梯度强度比较单元1231，用于将所述滤波后图片中当前像素点的梯度强度与沿正负梯度方向上的两个像素点进行比较，以判断当前像素点的梯度强度是否均大于当前像素点沿正负梯度方向上的两个像素点的梯度强度；其中，所述当前像素点(i，j)的初始值为(0,0)，所述滤波后图片中的总像素点个数为m*n个，其中m表示滤波后图片中横向像素点的总个数，n表示滤波后图片中纵向像素点的总个数，i的取值范围为[1，m-1]，j的取值范围为[1，n-1]，m、n均为大于1的自然数；
像素保留单元1232，用于若当前像素点的梯度强度大于当前像素点沿正负梯度方向上的两个像素点的梯度强度，将当前像素点的当前像素值进行保留；
像素抑制单元1233，用于若当前像素点的梯度强度小于当前像素点沿正负梯度方向上的两个像素点的梯度强度其中一个梯度强度，将当前像素点进行抑制，得到抑制后像素点；
像素点末位判断单元1234，用于判断当前像素点是否为所述滤波后图片中最后一个像素点；若当前像素点为所述滤波后图片中最后一个像素点，执行输出当前图片作为处理后图片的步骤；若当前像素点不为所述滤波后图片中最后一个像素点，获取与当前像素点向后相邻的下一像素点以更新作为当前像素点，返回执行将所述滤波后图片中当前像素点的梯度强度与沿正负梯度方向上的两个像素点进行比较，以判断当前像素点的梯度强度是否均大于当前像素点沿正负梯度方向上的两个像素点的梯度强度的步骤；
当前图片获取单元1235，用于输出当前图片作为处理后图片。
在本实施例中，即针对所述滤波后图片中各像素点进行非极大值抑制，是为了更加精确的获取图像的边缘点。具体的，在跨越梯度方向的两个相邻像素之间可使用线性插值来得到要比较的像素梯度，即若当前像素点的梯度强度大于当前像素点沿正负梯度方向上的两个像素点的梯度强度，将当前像素点的当前像素值进行保留；若当前像素点的梯度强度小于当前像素点沿正负梯度方向上的两个像素点的梯度强度其中一个梯度强度，将当前像素点进行抑制，得到抑制后像素点。通过非极大值抑制，则可以帮助将局部最大值之外的所有梯度值抑制为0，从而更准确的进行边缘检测。
当前间距值获取单元130，用于根据所述头部长度或是头部宽度与标准摄像头矩形框的对应边长之比以及标准摄像头矩形框的标准间距值，以获取用户与屏幕之间的当前间距值。
在本实施例中，由于智能终端的摄像头在采集用户的面部图像时，可以虚拟出面部检测的标准摄像头矩形框，该标准摄像头矩形框是固定不变的，不会随着用户与摄像头之间间距的变化而改变尺寸大小，通过所述当前用户人脸图像对应的头部长度或是头部宽度与该标准摄像头矩形框对应的标准长度或标准宽度的比例，即可估算出用户与屏幕之间的距离。
在一实施例中，当前间距值获取单元130包括：
标准摄像头矩形框获取单元，用于获取预先存储的标准摄像头矩形框，及所述标准摄像头矩形框对应的标准长度或标准宽度；
距离转化单元，用于根据所述头部长度与所述标准宽度之比、及标准摄像头矩形框对应的标准间距值，获取用户与屏幕之间的当前间距值；或者根据所述头部宽度与所述标准长度之比、及标准摄像头矩形框对应的标准间距值，获取用户与屏幕之间的当前间距值。
即通过所述当前的用户人脸图像对应的头部长度与该标准摄像头矩形框对应的标准宽度之比获取用户与屏幕之间的距离，或是通过所述当前的用户人脸图像对应的头部宽度与该标准摄像头矩形框对应的标准长度度之比获取用户与屏幕之间的距离。通过这一方式，能有效检测出用户与屏幕之间的间距从而估算视力值。
当前用户视力值获取单元140，用于根据当前间距值，及调用预先存储的间距值与视力映射关系表，以获取与当前间距值对应的当前用户视力值。
在本实施例中，可根据预设的用户与屏幕之间的距离值与视力值的换算曲线(该换算曲线可以理解为间距值与视力映射关系表)来进行换算，也即在间距值与视力映射关系表中每一距离值都对应一个视力值，这些间距值与视力映射关系表中距离值与视力值的对应关系是通过多次试验得到的有效数据。例如用户与屏幕之间的距离为30cm对应5.0的视力，用户与屏幕之间的距离为32cm对应5.1的视力等。通过这一方式，可以有效的将用户与屏幕之间的距离值换算成视力值，从而进行视力监测。
视力曲线构建单元150，用于根据所述当前用户视力值及已存储的用户历史视力值集合中满足预设的筛选条件的历史视力值集合，对应构建视力曲线；其中，所述视力曲线中以时间轴为X轴，且以各时刻对应的视力值为Y轴；所述筛选条件包括筛选时间段及视力值取值。
在本实施例中，可以在服务器中已存储的用户历史视力值集合中根据筛选条件，例如该筛选条件为筛选时间为2018年1月1日-2018年1月5日且不为空值的视力值，其中为空值的视力值是用户在当前时刻未使用智能终端时在视力曲线对应的时刻点上的体现。即若某一时刻视力值为空值，则不用绘制于该视力曲线图。通过绘制视力曲线，能有效监控用户的视力值变化趋势，从而可以获知是否有异常情况发生，也可以统计用户使用智能终端的时长。
曲线发送单元160，用于将所述视力曲线发送至用户端以进行显示。
在本实施例中，当在服务器中完成了根据预设的筛选条件而绘制的视力曲线后，可由服务器将所述视力曲线发送至用户端，以在用户端上的用户交互界面上进行直观展示，从而可以分析视力变化趋势。
在一实施例中，基于图像识别的视力检测装置100还包括：
异常区间获取单元，用于获取所述视力曲线中在单位时间内视力值变化超出预设的视力阈值的区间，以组成视力值异常变化区间；
高亮提示单元，用于将所述视力值异常变化区间进行高亮显示。
在本实施例中，若在所述视力曲线中检测到有某一时刻的视力值与前一时刻的视力值及后一时刻的视力值均发生显著变化，例如当前时刻的视力值为4.5，其前一时刻的视力值为5.0，其后一时刻的视力值为5.0，表示当前时刻的视力值可能存在异常，需要对用户进行提示以重视视力保护。此时，可将所述视力值异常变化区间在视力曲线上进行高亮显示(如红色)，通过这一直观的方式能对客户进行有效提示。
该装置实现了使用带有前置摄像头的智能终端时自动周期性采集用户人脸图像以检测用户视力，还可自动生成视力曲线，不仅提高了检测效率，而且获取了海量的检测数据以供数据分析使用。
上述基于图像识别的视力检测装置可以实现为计算机程序的形式，该计算机程序可以在如图10所示的计算机设备上运行。
请参阅图10，图10是本申请实施例提供的计算机设备的示意性框图。该计算机设备500是服务器，服务器可以是独立的服务器，也可以是多个服务器组成的服务器集群。
参阅图10，该计算机设备500包括通过系统总线501连接的处理器502、存储器和网络接口505，其中，存储器可以包括非易失性存储介质503和内存储器504。
该非易失性存储介质503可存储操作系统5031和计算机程序5032。该计算机程序5032 被执行时，可使得处理器502执行基于图像识别的视力检测方法。
该处理器502用于提供计算和控制能力，支撑整个计算机设备500的运行。
该内存储器504为非易失性存储介质503中的计算机程序5032的运行提供环境，该计算机程序5032被处理器502执行时，可使得处理器502执行基于图像识别的视力检测方法。
该网络接口505用于进行网络通信，如提供数据信息的传输等。本领域技术人员可以理解，图10中示出的结构，仅仅是与本申请方案相关的部分结构的框图，并不构成对本申请方案所应用于其上的计算机设备500的限定，具体的计算机设备500可以包括比图中所示更多或更少的部件，或者组合某些部件，或者具有不同的部件布置。
其中，所述处理器502用于运行存储在存储器中的计算机程序5032，以实现本申请实施例公开的基于图像识别的视力检测方法。
本领域技术人员可以理解，图10中示出的计算机设备的实施例并不构成对计算机设备具体构成的限定，在其他实施例中，计算机设备可以包括比图示更多或更少的部件，或者组合某些部件，或者不同的部件布置。例如，在一些实施例中，计算机设备可以仅包括存储器及处理器，在这样的实施例中，存储器及处理器的结构及功能与图10所示实施例一致，在此不再赘述。
应当理解，在本申请实施例中，处理器502可以是中央处理单元(Central ProcessingUnit，CPU)，该处理器502还可以是其他通用处理器、数字信号处理器(Digital Signal Processor，DSP)、专用集成电路(Application Specific Integrated Circuit，ASIC)、现成可编程门阵列(Field-Programmable GateArray，FPGA)或者其他可编程逻辑器件、分立门或者晶体管逻辑器件、分立硬件组件等。其中，通用处理器可以是微处理器或者该处理器也可以是任何常规的处理器等。
在本申请的另一实施例中提供计算机可读存储介质。该计算机可读存储介质可以为非易失性的计算机可读存储介质，也可以为易失性的计算机可读存储介质。该计算机可读存储介质存储有计算机程序，其中计算机程序被处理器执行时实现本申请实施例公开的基于图像识别的视力检测方法。
所属领域的技术人员可以清楚地了解到，为了描述的方便和简洁，上述描述的设备、装置和单元的具体工作过程，可以参考前述方法实施例中的对应过程，在此不再赘述。本领域普通技术人员可以意识到，结合本文中所公开的实施例描述的各示例的单元及算法步骤，能够以电子硬件、计算机软件或者二者的结合来实现，为了清楚地说明硬件和软件的可互换性，在上述说明中已经按照功能一般性地描述了各示例的组成及步骤。这些功能究竟以硬件还是软件方式来执行取决于技术方案的特定应用和设计约束条件。专业技术人员可以对每个特定的应用来使用不同方法来实现所描述的功能，但是这种实现不应认为超出本申请的范围。
在本申请所提供的几个实施例中，应该理解到，所揭露的设备、装置和方法，可以通过其它的方式实现。例如，以上所描述的装置实施例仅仅是示意性的，例如，所述单元的划分，仅仅为逻辑功能划分，实际实现时可以有另外的划分方式，也可以将具有相同功能的单元集合成一个单元，例如多个单元或组件可以结合或者可以集成到另一个系统，或一些特征可以忽略，或不执行。另外，所显示或讨论的相互之间的耦合或直接耦合或通信连接可以是通过一些接口、装置或单元的间接耦合或通信连接，也可以是电的，机械的或其它的形式连接。
所述作为分离部件说明的单元可以是或者也可以不是物理上分开的，作为单元显示的部件可以是或者也可以不是物理单元，即可以位于一个地方，或者也可以分布到多个网络单元上。可以根据实际的需要选择其中的部分或者全部单元来实现本申请实施例方案的目的。
另外，在本申请各个实施例中的各功能单元可以集成在一个处理单元中，也可以是各个单元单独物理存在，也可以是两个或两个以上单元集成在一个单元中。上述集成的单元既可以采用硬件的形式实现，也可以采用软件功能单元的形式实现。
所述集成的单元如果以软件功能单元的形式实现并作为独立的产品销售或使用时，可以存储在一个存储介质中。基于这样的理解，本申请的技术方案本质上或者说对现有技术做出 贡献的部分，或者该技术方案的全部或部分可以以软件产品的形式体现出来，该计算机软件产品存储在一个存储介质中，包括若干指令用以使得一台计算机设备(可以是个人计算机，服务器，或者网络设备等)执行本申请各个实施例所述方法的全部或部分步骤。而前述的存储介质包括：U盘、移动硬盘、只读存储器(ROM，Read-Only Memory)、磁碟或者光盘等各种可以存储程序代码的介质。
以上所述，仅为本申请的具体实施方式，但本申请的保护范围并不局限于此，任何熟悉本技术领域的技术人员在本申请揭露的技术范围内，可轻易想到各种等效的修改或替换，这些修改或替换都应涵盖在本申请的保护范围之内。因此，本申请的保护范围应以权利要求的保护范围为准。

Claims (20)

  1. 一种基于图像识别的视力检测方法，其中，包括：
    若当前系统时刻与上一图片采集时刻之差等于预设的图片采集周期，接收用户端所上传的当前用户人脸图像；
    获取当前用户人脸图像中的头部长度或是头部宽度；
    根据所述头部长度或是头部宽度与标准摄像头矩形框的对应边长之比以及标准摄像头矩形框的标准间距值，以获取用户与屏幕之间的当前间距值；
    根据当前间距值，及调用预先存储的间距值与视力映射关系表，以获取与当前间距值对应的当前用户视力值；
    根据所述当前用户视力值及已存储的用户历史视力值集合中满足预设的筛选条件的历史视力值集合，对应构建视力曲线；其中，所述视力曲线中以时间轴为X轴，且以各时刻对应的视力值为Y轴；所述筛选条件包括筛选时间段及视力值取值；以及
    将所述视力曲线发送至用户端以进行显示。
  2. 根据权利要求1所述的基于图像识别的视力检测方法，其中，所述获取当前用户人脸图像中的头部长度或是头部宽度，包括：
    将所述当前用户人脸图像进行灰度化，得到灰度化图片；
    将所述灰度化图片进行高斯滤波，得到滤波后图片；
    获取所述滤波后图片的梯度值和方向，并将所述滤波后图片进行非极大值抑制，以得到处理后图片；
    将所述处理后图片进行双阈值检测及连接边缘，得到与所述当前的用户人脸图像对应的当前人脸边缘检测图像；
    根据所述当前人脸边缘检测图像获取对应的当前人脸矩形框，以根据所述当前人脸矩形框获取所述当前的用户人脸图像对应的头部长度或是头部宽度。
  3. 根据权利要求2所述的基于图像识别的视力检测方法，其中，所述将所述滤波后图片进行非极大值抑制以得到处理后图片，包括：
    将所述滤波后图片中当前像素点的梯度强度与沿正负梯度方向上的两个像素点进行比较，以判断当前像素点的梯度强度是否均大于当前像素点沿正负梯度方向上的两个像素点的梯度强度；其中，所述当前像素点(i，j)的初始值为(0,0)，所述滤波后图片中的总像素点个数为m*n个，其中m表示滤波后图片中横向像素点的总个数，n表示滤波后图片中纵向像素点的总个数，i的取值范围为[1，m-1]，j的取值范围为[1，n-1]，m、n均为大于1的自然数；
    若当前像素点的梯度强度大于当前像素点沿正负梯度方向上的两个像素点的梯度强度，将当前像素点的当前像素值进行保留；
    若当前像素点的梯度强度小于当前像素点沿正负梯度方向上的两个像素点的梯度强度其中一个梯度强度，将当前像素点进行抑制，得到抑制后像素点；
    判断当前像素点是否为所述滤波后图片中最后一个像素点；若当前像素点为所述滤波后图片中最后一个像素点，执行输出当前图片作为处理后图片的步骤；若当前像素点不为所述滤波后图片中最后一个像素点，获取与当前像素点向后相邻的下一像素点以更新作为当前像素点，返回执行将所述滤波后图片中当前像素点的梯度强度与沿正负梯度方向上的两个像素点进行比较，以判断当前像素点的梯度强度是否均大于当前像素点沿正负梯度方向上的两个像素点的梯度强度的步骤；
    输出当前图片作为处理后图片。
  4. 根据权利要求1所述的基于图像识别的视力检测方法，其中，所述将所述当前用户人脸图像进行灰度化，得到灰度化图片，包括：
    获取所述当前用户人脸图像中每个像素点(I，J)对应的R通道值、G通道值、B通道 值；其中，当前用户人脸图像片的总像素点个数为M*N个，其中M表示当前用户人脸图像片中横向像素点的总个数，N表示当前用户人脸图像片中纵向像素点的总个数，I的取值范围为[0，M-1]，j的取值范围为[0，N-1]；
    根据所述当前用户人脸图像中每个像素点(I，J)对应的R通道值R IJ、G通道值G IJ、B通道值B IJ及Gray IJ＝R IJ*0.299+G IJ*0.587+B IJ*0.114，对应计算获取每个像素点(I，J)对应的灰度值Gray IJ；其中，Gray IJ表示像素点(I，J)的灰度值，R IJ表示像素点(I，J)对应的R通道值、G IJ表示像素点(I，J)对应的G通道值、B IJ表示像素点(I，J)对应的B通道值；
    将所述当前用户人脸图像中每个像素点(I，J)对应的R通道值、G通道值、B通道值转化为对应的灰度值Gray ij，以得到对应的灰度化图片。
  5. 根据权利要求1所述的基于图像识别的视力检测方法，其中，所述根据所述头部长度或是头部宽度与标准摄像头矩形框的对应边长之比以及标准摄像头矩形框的标准间距值，以获取用户与屏幕之间的当前间距值，包括：
    获取预先存储的标准摄像头矩形框，及所述标准摄像头矩形框对应的标准长度或标准宽度；
    根据所述头部长度与所述标准宽度之比、及标准摄像头矩形框对应的标准间距值，获取用户与屏幕之间的当前间距值；或者根据所述头部宽度与所述标准长度之比、及标准摄像头矩形框对应的标准间距值，获取用户与屏幕之间的当前间距值。
  6. 根据权利要求1所述的基于图像识别的视力检测方法，其中，所述将所述视力曲线发送至用户端以进行显示之后，还包括：
    获取所述视力曲线中在单位时间内视力值变化超出预设的视力阈值的区间，以组成视力值异常变化区间；
    将所述视力值异常变化区间进行高亮显示。
  7. 一种基于图像识别的视力检测装置，其中，包括：
    接收单元，用于若当前系统时刻与上一图片采集时刻之差等于预设的图片采集周期，接收用户端所上传的当前用户人脸图像；
    尺寸获取单元，用于获取当前用户人脸图像中的头部长度或是头部宽度；
    当前间距值获取单元，用于根据所述头部长度或是头部宽度与标准摄像头矩形框的对应边长之比以及标准摄像头矩形框的标准间距值，以获取用户与屏幕之间的当前间距值；
    当前用户视力值获取单元，用于根据当前间距值，及调用预先存储的间距值与视力映射关系表，以获取与当前间距值对应的当前用户视力值；
    视力曲线构建单元，用于根据所述当前用户视力值及已存储的用户历史视力值集合中满足预设的筛选条件的历史视力值集合，对应构建视力曲线；其中，所述视力曲线中以时间轴为X轴，且以各时刻对应的视力值为Y轴；所述筛选条件包括筛选时间段及视力值取值；以及
    曲线发送单元，用于将所述视力曲线发送至用户端以进行显示。
  8. 根据权利要求7所述的基于图像识别的视力检测装置，其中，所述尺寸获取单元，包括：
    灰度化单元，用于将所述当前用户人脸图像进行灰度化，得到灰度化图片；
    滤波单元，用于将所述灰度化图片进行高斯滤波，得到滤波后图片；
    非极值抑制单元，用于获取所述滤波后图片的梯度值和方向，并将所述滤波后图片进行非极大值抑制，以得到处理后图片；
    边缘检测单元，用于将所述处理后图片进行双阈值检测及连接边缘，得到与所述当前的用户人脸图像对应的当前人脸边缘检测图像；
    当前人脸矩形框获取单元，用于根据所述当前人脸边缘检测图像获取对应的当前人脸矩形框，以根据所述当前人脸矩形框获取所述当前的用户人脸图像对应的头部长度或是头部宽度。
  9. 一种计算机设备，包括存储器、处理器及存储在所述存储器上并可在所述处理器上运行的计算机程序，其中，所述处理器执行所述计算机程序时实现如下步骤：
    若当前系统时刻与上一图片采集时刻之差等于预设的图片采集周期，接收用户端所上传的当前用户人脸图像；
    获取当前用户人脸图像中的头部长度或是头部宽度；
    根据所述头部长度或是头部宽度与标准摄像头矩形框的对应边长之比以及标准摄像头矩形框的标准间距值，以获取用户与屏幕之间的当前间距值；
    根据当前间距值，及调用预先存储的间距值与视力映射关系表，以获取与当前间距值对应的当前用户视力值；
    根据所述当前用户视力值及已存储的用户历史视力值集合中满足预设的筛选条件的历史视力值集合，对应构建视力曲线；其中，所述视力曲线中以时间轴为X轴，且以各时刻对应的视力值为Y轴；所述筛选条件包括筛选时间段及视力值取值；以及
    将所述视力曲线发送至用户端以进行显示。
  10. 如权利要求9所述的计算机设备，其中，所述处理器执行所述计算机程序时实现所述获取当前用户人脸图像中的头部长度或是头部宽度的步骤，包括：
    将所述当前用户人脸图像进行灰度化，得到灰度化图片；
    将所述灰度化图片进行高斯滤波，得到滤波后图片；
    获取所述滤波后图片的梯度值和方向，并将所述滤波后图片进行非极大值抑制，以得到处理后图片；
    将所述处理后图片进行双阈值检测及连接边缘，得到与所述当前的用户人脸图像对应的当前人脸边缘检测图像；
    根据所述当前人脸边缘检测图像获取对应的当前人脸矩形框，以根据所述当前人脸矩形框获取所述当前的用户人脸图像对应的头部长度或是头部宽度。
  11. 如权利要求10所述的计算机设备，其中，所述处理器执行所述计算机程序时实现所述将所述滤波后图片进行非极大值抑制以得到处理后图片的步骤，包括：
    将所述滤波后图片中当前像素点的梯度强度与沿正负梯度方向上的两个像素点进行比较，以判断当前像素点的梯度强度是否均大于当前像素点沿正负梯度方向上的两个像素点的梯度强度；其中，所述当前像素点(i，j)的初始值为(0,0)，所述滤波后图片中的总像素点个数为m*n个，其中m表示滤波后图片中横向像素点的总个数，n表示滤波后图片中纵向像素点的总个数，i的取值范围为[1，m-1]，j的取值范围为[1，n-1]，m、n均为大于1的自然数；
    若当前像素点的梯度强度大于当前像素点沿正负梯度方向上的两个像素点的梯度强度，将当前像素点的当前像素值进行保留；
    若当前像素点的梯度强度小于当前像素点沿正负梯度方向上的两个像素点的梯度强度其中一个梯度强度，将当前像素点进行抑制，得到抑制后像素点；
    判断当前像素点是否为所述滤波后图片中最后一个像素点；若当前像素点为所述滤波后图片中最后一个像素点，执行输出当前图片作为处理后图片的步骤；若当前像素点不为所述滤波后图片中最后一个像素点，获取与当前像素点向后相邻的下一像素点以更新作为当前像素点，返回执行将所述滤波后图片中当前像素点的梯度强度与沿正负梯度方向上的两个像素点进行比较，以判断当前像素点的梯度强度是否均大于当前像素点沿正负梯度方向上的两个像素点的梯度强度的步骤；
    输出当前图片作为处理后图片。
  12. 如权利要求9所述的计算机设备，其中，所述处理器执行所述计算机程序时实现所述将所述当前用户人脸图像进行灰度化，得到灰度化图片的步骤，包括：
    获取所述当前用户人脸图像中每个像素点(I，J)对应的R通道值、G通道值、B通道值；其中，当前用户人脸图像片的总像素点个数为M*N个，其中M表示当前用户人脸图像 片中横向像素点的总个数，N表示当前用户人脸图像片中纵向像素点的总个数，I的取值范围为[0，M-1]，j的取值范围为[0，N-1]；
    根据所述当前用户人脸图像中每个像素点(I，J)对应的R通道值R IJ、G通道值G IJ、B通道值B IJ及Gray IJ＝R IJ*0.299+G IJ*0.587+B IJ*0.114，对应计算获取每个像素点(I，J)对应的灰度值Gray IJ；其中，Gray IJ表示像素点(I，J)的灰度值，R IJ表示像素点(I，J)对应的R通道值、G IJ表示像素点(I，J)对应的G通道值、B IJ表示像素点(I，J)对应的B通道值；
    将所述当前用户人脸图像中每个像素点(I，J)对应的R通道值、G通道值、B通道值转化为对应的灰度值Gray ij，以得到对应的灰度化图片。
  13. 如权利要求9所述的计算机设备，其中，所述处理器执行所述计算机程序时实现所述根据所述头部长度或是头部宽度与标准摄像头矩形框的对应边长之比以及标准摄像头矩形框的标准间距值，以获取用户与屏幕之间的当前间距值的步骤，包括：
    获取预先存储的标准摄像头矩形框，及所述标准摄像头矩形框对应的标准长度或标准宽度；
    根据所述头部长度与所述标准宽度之比、及标准摄像头矩形框对应的标准间距值，获取用户与屏幕之间的当前间距值；或者根据所述头部宽度与所述标准长度之比、及标准摄像头矩形框对应的标准间距值，获取用户与屏幕之间的当前间距值。
  14. 如权利要求9所述的计算机设备，其中，所述处理器执行所述计算机程序时实现所述将所述视力曲线发送至用户端以进行显示的步骤之后，还用于实现如下步骤：
    获取所述视力曲线中在单位时间内视力值变化超出预设的视力阈值的区间，以组成视力值异常变化区间；
    将所述视力值异常变化区间进行高亮显示。
  15. 一种计算机可读存储介质，其中，所述计算机可读存储介质存储有计算机程序，所述计算机程序当被处理器执行时使所述处理器执行如下步骤：
    若当前系统时刻与上一图片采集时刻之差等于预设的图片采集周期，接收用户端所上传的当前用户人脸图像；
    获取当前用户人脸图像中的头部长度或是头部宽度；
    根据所述头部长度或是头部宽度与标准摄像头矩形框的对应边长之比以及标准摄像头矩形框的标准间距值，以获取用户与屏幕之间的当前间距值；
    根据当前间距值，及调用预先存储的间距值与视力映射关系表，以获取与当前间距值对应的当前用户视力值；
    根据所述当前用户视力值及已存储的用户历史视力值集合中满足预设的筛选条件的历史视力值集合，对应构建视力曲线；其中，所述视力曲线中以时间轴为X轴，且以各时刻对应的视力值为Y轴；所述筛选条件包括筛选时间段及视力值取值；以及
    将所述视力曲线发送至用户端以进行显示。
  16. 如权利要求15所述的存储介质，其中，所述计算机程序当被处理器执行时使所述处理器执行所述获取当前用户人脸图像中的头部长度或是头部宽度的步骤，包括：
    将所述当前用户人脸图像进行灰度化，得到灰度化图片；
    将所述灰度化图片进行高斯滤波，得到滤波后图片；
    获取所述滤波后图片的梯度值和方向，并将所述滤波后图片进行非极大值抑制，以得到处理后图片；
    将所述处理后图片进行双阈值检测及连接边缘，得到与所述当前的用户人脸图像对应的当前人脸边缘检测图像；
    根据所述当前人脸边缘检测图像获取对应的当前人脸矩形框，以根据所述当前人脸矩形框获取所述当前的用户人脸图像对应的头部长度或是头部宽度。
  17. 如权利要求16所述的存储介质，其中，所述计算机程序当被处理器执行时使所述处理器执行所述将所述滤波后图片进行非极大值抑制以得到处理后图片的步骤，包括：将所述 滤波后图片中当前像素点的梯度强度与沿正负梯度方向上的两个像素点进行比较，以判断当前像素点的梯度强度是否均大于当前像素点沿正负梯度方向上的两个像素点的梯度强度；其中，所述当前像素点(i，j)的初始值为(0,0)，所述滤波后图片中的总像素点个数为m*n个，其中m表示滤波后图片中横向像素点的总个数，n表示滤波后图片中纵向像素点的总个数，i的取值范围为[1，m-1]，j的取值范围为[1，n-1]，m、n均为大于1的自然数；
    若当前像素点的梯度强度大于当前像素点沿正负梯度方向上的两个像素点的梯度强度，将当前像素点的当前像素值进行保留；
    若当前像素点的梯度强度小于当前像素点沿正负梯度方向上的两个像素点的梯度强度其中一个梯度强度，将当前像素点进行抑制，得到抑制后像素点；
    判断当前像素点是否为所述滤波后图片中最后一个像素点；若当前像素点为所述滤波后图片中最后一个像素点，执行输出当前图片作为处理后图片的步骤；若当前像素点不为所述滤波后图片中最后一个像素点，获取与当前像素点向后相邻的下一像素点以更新作为当前像素点，返回执行将所述滤波后图片中当前像素点的梯度强度与沿正负梯度方向上的两个像素点进行比较，以判断当前像素点的梯度强度是否均大于当前像素点沿正负梯度方向上的两个像素点的梯度强度的步骤；
    输出当前图片作为处理后图片。
  18. 如权利要求15所述的存储介质，其中，所述计算机程序当被处理器执行时使所述处理器执行所述将所述当前用户人脸图像进行灰度化，得到灰度化图片的步骤，包括：
    获取所述当前用户人脸图像中每个像素点(I，J)对应的R通道值、G通道值、B通道值；其中，当前用户人脸图像片的总像素点个数为M*N个，其中M表示当前用户人脸图像片中横向像素点的总个数，N表示当前用户人脸图像片中纵向像素点的总个数，I的取值范围为[0，M-1]，j的取值范围为[0，N-1]；
    根据所述当前用户人脸图像中每个像素点(I，J)对应的R通道值R IJ、G通道值G IJ、B通道值B IJ及Gray IJ＝R IJ*0.299+G IJ*0.587+B IJ*0.114，对应计算获取每个像素点(I，J)对应的灰度值Gray IJ；其中，Gray IJ表示像素点(I，J)的灰度值，R IJ表示像素点(I，J)对应的R通道值、G IJ表示像素点(I，J)对应的G通道值、B IJ表示像素点(I，J)对应的B通道值；
    将所述当前用户人脸图像中每个像素点(I，J)对应的R通道值、G通道值、B通道值转化为对应的灰度值Gray ij，以得到对应的灰度化图片。
  19. 如权利要求15所述的存储介质，其中，所述计算机程序当被处理器执行时使所述处理器执行所述根据所述头部长度或是头部宽度与标准摄像头矩形框的对应边长之比以及标准摄像头矩形框的标准间距值，以获取用户与屏幕之间的当前间距值的步骤，包括：
    获取预先存储的标准摄像头矩形框，及所述标准摄像头矩形框对应的标准长度或标准宽度；
    根据所述头部长度与所述标准宽度之比、及标准摄像头矩形框对应的标准间距值，获取用户与屏幕之间的当前间距值；或者根据所述头部宽度与所述标准长度之比、及标准摄像头矩形框对应的标准间距值，获取用户与屏幕之间的当前间距值。
  20. 如权利要求15所述的存储介质，其中，所述计算机程序当被处理器执行时使所述处理器执行所述将所述视力曲线发送至用户端以进行显示的步骤之后，还执行如下步骤：
    获取所述视力曲线中在单位时间内视力值变化超出预设的视力阈值的区间，以组成视力值异常变化区间；
    将所述视力值异常变化区间进行高亮显示。
PCT/CN2020/087032 2019-10-12 2020-04-26 基于图像识别的视力检测方法、装置、及计算机设备 WO2021068486A1 (zh)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN201910969805.XA CN111084606A (zh) 2019-10-12 2019-10-12 基于图像识别的视力检测方法、装置、及计算机设备
CN201910969805.X 2019-10-12

Publications (1)

Publication Number Publication Date
WO2021068486A1 true WO2021068486A1 (zh) 2021-04-15

Family

ID=70392968

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2020/087032 WO2021068486A1 (zh) 2019-10-12 2020-04-26 基于图像识别的视力检测方法、装置、及计算机设备

Country Status (2)

Country Link
CN (1) CN111084606A (zh)
WO (1) WO2021068486A1 (zh)

Cited By (9)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN114331991A (zh) * 2021-12-21 2022-04-12 国能朔黄铁路发展有限责任公司 小半径曲线无缝线路监测方法、系统、装置和计算机设备
CN114463326A (zh) * 2022-03-14 2022-05-10 深圳灿维科技有限公司 一种手机中框视觉检测算法、装置、设备及存储介质
CN115029704A (zh) * 2022-08-09 2022-09-09 江苏贯森新材料科技有限公司 一种用于不锈钢酸洗工艺的智能控制方法及装置
CN115187552A (zh) * 2022-07-14 2022-10-14 南京思飞捷软件科技有限公司 一种基于边缘检测的车身门缝检测方法
WO2023103088A1 (zh) * 2021-12-09 2023-06-15 深圳创维－Rgb电子有限公司 视力检测方法、电子设备以及存储介质
CN116385706A (zh) * 2023-06-06 2023-07-04 山东外事职业大学 基于图像识别技术的信号检测方法和系统
CN116485884A (zh) * 2023-06-28 2023-07-25 四川君安天源精酿啤酒有限公司 基于计算机视觉的精酿啤酒瓶口实时定位方法及系统
CN116630321A (zh) * 2023-07-24 2023-08-22 铁正检测科技有限公司 基于人工智能的桥梁健康智能监测系统
CN117455908A (zh) * 2023-12-22 2024-01-26 山东济矿鲁能煤电股份有限公司阳城煤矿 一种皮带机跑偏视觉检测方法及系统

Families Citing this family (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
WO2021248671A1 (zh) * 2020-06-12 2021-12-16 海信视像科技股份有限公司 一种显示设备
CN113344999A (zh) * 2021-06-28 2021-09-03 北京市商汤科技开发有限公司 深度检测方法及装置、电子设备和存储介质
CN113674272B (zh) * 2021-09-06 2024-03-15 上海集成电路装备材料产业创新中心有限公司 图像检测方法与装置
CN115732091A (zh) * 2022-11-24 2023-03-03 深圳安视信息技术有限公司 视力自助筛查系统

Citations (9)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20140293228A1 (en) * 2009-08-17 2014-10-02 Allan N. Hytowitz Visual Acuity-Based Moving Image
CN106491075A (zh) * 2016-11-29 2017-03-15 珠海格力电器股份有限公司 一种视力检测方法、装置和视力检测器
CN106725288A (zh) * 2016-11-29 2017-05-31 珠海格力电器股份有限公司 视力检测方法、装置和移动终端
CN107198505A (zh) * 2017-04-07 2017-09-26 天津市天中依脉科技开发有限公司 基于智能手机的视功能检测系统及方法
CN108652579A (zh) * 2018-04-25 2018-10-16 珠海格力电器股份有限公司 一种检测视力的方法和终端
CN109330556A (zh) * 2018-09-29 2019-02-15 杭州艾斯凯尔科技有限公司 利用图像技术和云端控制技术的视力检测方法和系统
CN110123257A (zh) * 2019-03-29 2019-08-16 深圳和而泰家居在线网络科技有限公司 一种视力检测方法、装置、视力检测仪及计算机存储介质
CN110279391A (zh) * 2019-05-30 2019-09-27 廖亮举 便携红外相机视力检测算法
JP2019169080A (ja) * 2018-03-26 2019-10-03 株式会社Ｊｖｃケンウッド 視線検出装置、視線検出方法及び視線検出プログラム

Family Cites Families (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN101533327A (zh) * 2008-03-13 2009-09-16 英业达股份有限公司 窗口调整界面及其调整方法
RU2706372C9 (ru) * 2013-06-06 2020-08-12 6 ОВЕР 6 ВИЖН Лтд Система и способ измерения рефракционной ошибки глаза на основании субъективного измерения расстояния
CN106022209B (zh) * 2016-04-29 2019-09-17 杭州华橙网络科技有限公司 一种基于人脸检测的距离估算和处理的方法及装置
EP3295863B1 (en) * 2016-09-15 2023-04-19 Essilor International Measurement method for determining a value of a visual correction need for near vision of an individual in a natural posture for near vision
CN106548135A (zh) * 2016-10-17 2017-03-29 北海益生源农贸有限责任公司 一种路面障碍检测方法
CN109326339A (zh) * 2018-12-07 2019-02-12 北京大学第三医院 一种视功能检查建议确定方法、装置、设备和介质
CN109464118B (zh) * 2018-12-10 2021-07-27 西安米克跨界信息科技有限公司 一种基于手机视力检测和眼镜配制方法

Patent Citations (9)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20140293228A1 (en) * 2009-08-17 2014-10-02 Allan N. Hytowitz Visual Acuity-Based Moving Image
CN106491075A (zh) * 2016-11-29 2017-03-15 珠海格力电器股份有限公司 一种视力检测方法、装置和视力检测器
CN106725288A (zh) * 2016-11-29 2017-05-31 珠海格力电器股份有限公司 视力检测方法、装置和移动终端
CN107198505A (zh) * 2017-04-07 2017-09-26 天津市天中依脉科技开发有限公司 基于智能手机的视功能检测系统及方法
JP2019169080A (ja) * 2018-03-26 2019-10-03 株式会社Ｊｖｃケンウッド 視線検出装置、視線検出方法及び視線検出プログラム
CN108652579A (zh) * 2018-04-25 2018-10-16 珠海格力电器股份有限公司 一种检测视力的方法和终端
CN109330556A (zh) * 2018-09-29 2019-02-15 杭州艾斯凯尔科技有限公司 利用图像技术和云端控制技术的视力检测方法和系统
CN110123257A (zh) * 2019-03-29 2019-08-16 深圳和而泰家居在线网络科技有限公司 一种视力检测方法、装置、视力检测仪及计算机存储介质
CN110279391A (zh) * 2019-05-30 2019-09-27 廖亮举 便携红外相机视力检测算法

Cited By (16)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
WO2023103088A1 (zh) * 2021-12-09 2023-06-15 深圳创维－Rgb电子有限公司 视力检测方法、电子设备以及存储介质
CN114331991A (zh) * 2021-12-21 2022-04-12 国能朔黄铁路发展有限责任公司 小半径曲线无缝线路监测方法、系统、装置和计算机设备
CN114331991B (zh) * 2021-12-21 2024-03-15 国能朔黄铁路发展有限责任公司 小半径曲线无缝线路监测方法、系统、装置和计算机设备
CN114463326A (zh) * 2022-03-14 2022-05-10 深圳灿维科技有限公司 一种手机中框视觉检测算法、装置、设备及存储介质
CN114463326B (zh) * 2022-03-14 2022-06-21 深圳灿维科技有限公司 一种手机中框视觉检测算法、装置、设备及存储介质
CN115187552A (zh) * 2022-07-14 2022-10-14 南京思飞捷软件科技有限公司 一种基于边缘检测的车身门缝检测方法
CN115029704A (zh) * 2022-08-09 2022-09-09 江苏贯森新材料科技有限公司 一种用于不锈钢酸洗工艺的智能控制方法及装置
CN115029704B (zh) * 2022-08-09 2022-12-09 江苏贯森新材料科技有限公司 一种用于不锈钢酸洗工艺的智能控制方法及装置
CN116385706B (zh) * 2023-06-06 2023-08-25 山东外事职业大学 基于图像识别技术的信号检测方法和系统
CN116385706A (zh) * 2023-06-06 2023-07-04 山东外事职业大学 基于图像识别技术的信号检测方法和系统
CN116485884A (zh) * 2023-06-28 2023-07-25 四川君安天源精酿啤酒有限公司 基于计算机视觉的精酿啤酒瓶口实时定位方法及系统
CN116485884B (zh) * 2023-06-28 2023-09-12 四川君安天源精酿啤酒有限公司 基于计算机视觉的精酿啤酒瓶口实时定位方法及系统
CN116630321B (zh) * 2023-07-24 2023-10-03 铁正检测科技有限公司 基于人工智能的桥梁健康智能监测系统
CN116630321A (zh) * 2023-07-24 2023-08-22 铁正检测科技有限公司 基于人工智能的桥梁健康智能监测系统
CN117455908A (zh) * 2023-12-22 2024-01-26 山东济矿鲁能煤电股份有限公司阳城煤矿 一种皮带机跑偏视觉检测方法及系统
CN117455908B (zh) * 2023-12-22 2024-04-09 山东济矿鲁能煤电股份有限公司阳城煤矿 一种皮带机跑偏视觉检测方法及系统

Also Published As

Publication number Publication date
CN111084606A (zh) 2020-05-01

Similar Documents

Publication Publication Date Title
WO2021068486A1 (zh) 基于图像识别的视力检测方法、装置、及计算机设备
US9888186B2 (en) Image acquisition method and image acquisition apparatus
EP3192008B1 (en) Systems and methods for liveness analysis
US10321072B2 (en) Image obtaining apparatus and method for obtaining an image based on pixel binning
WO2019137038A1 (zh) 注视点确定方法、对比度调节方法及对比度调节装置、虚拟现实设备及存储介质
JP6615917B2 (ja) 実時間ビデオエンハンスメント方法、端末及び非一時的コンピュータ可読記憶媒体
CN110059666B (zh) 一种注意力检测方法及装置
CN104182721A (zh) 提升人脸识别率的图像处理系统及图像处理方法
CN111444555B (zh) 一种测温信息显示方法、装置及终端设备
US9147115B2 (en) Method and device for detecting an object in an image
CN104954627B (zh) 一种信息处理方法及电子设备
CN107464225B (zh) 图像处理方法、装置、计算机可读存储介质和移动终端
CN110321782B (zh) 一种检测人体特征信号的系统
Mikhailiuk et al. The effect of display brightness and viewing distance: a dataset for visually lossless image compression
WO2017143654A1 (zh) 选择待输出照片的方法、拍照方法、装置及存储介质
Bianco et al. Image quality assessment by preprocessing and full reference model combination
CN115022616A (zh) 一种基于人眼跟踪的图像聚焦增强显示装置和显示方法
CN109598276A (zh) 图像处理装置和方法及监控系统
Field et al. Does spatial invariance result from insensitivity to change?
Dauphin et al. Background suppression with low-resolution camera in the context of medication intake monitoring
CN115829887A (zh) 血管图像的处理方法、装置和电子设备
WO2019109061A1 (en) Temporal de-noising
Lee et al. Image quality assessment of tone mapped images
CN115118947B (zh) 图像的处理方法、装置、电子设备及存储介质
CN112614151B (zh) 运动事件检测方法、电子设备及计算机可读存储介质

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 20874905

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

32PN Ep: public notification in the ep bulletin as address of the adressee cannot be established

Free format text: NOTING OF LOSS OF RIGHTS PURSUANT TO RULE 112(1) EPC (EPO FORM 1205A DATED 18/08/2022)

122 Ep: pct application non-entry in european phase

Ref document number: 20874905

Country of ref document: EP

Kind code of ref document: A1