TWM480134U - Monitor device based on image recognition - Google Patents

Description

Image recognition based monitoring device

The present invention relates to a monitoring device, and more particularly to a monitoring device based on image recognition.

In general, a baby born to three months old will only swing his hands and feet, and the neck muscles are not strong enough to turn or lift; while with age, muscle strength increases, and the baby reaches four months old. When the head begins to lift slightly and rotate left and right, physical activity begins to appear at five or six months old, such as sideways flipping and creeping. At this time, it is more prone to suffocation or accidents falling under the bed. In addition, elderly people who have been bedridden for a long time are very similar to those of infants within six months of age.

It can be seen from the above that small infants and elderly people who have been bedridden for a long time are in great need of safety monitoring. A common type of security monitoring is human-powered care, but the way people look after it is not only labor-saving, but also difficult to achieve 24-hour continuous security monitoring, so we must find a solution.

Therefore, the object of the present invention is to provide a monitoring device.

Therefore, the novel image recognition based monitoring device is suitable for monitoring a user of a bed, and comprises an image capturing unit and a memory , a processing unit and a warning unit. The image capturing unit is configured to capture a continuous first image and a second image of the user, wherein the capturing of the second image occurs after the capturing of the first image.

The memory is configured to store a face detection program module and a motion detection program module, wherein the face detection program module is configured to detect a first face of the user in the first image And a second face block of the user in the second image, the motion detection program module is configured to detect whether the second face block is opposite to the first face block The degree of movement of the face exceeds a preset threshold.

The processing unit is configured to execute the face detection program module and the motion detection program module.

When the motion detection program module detects that the movement of the second facial block relative to the face of the first facial block exceeds the preset threshold, the processing unit triggers the warning unit to be outward Send a warning.

The effect of the present invention is that the user's face movement degree can be detected by the setting of the image capturing unit, the face detection program module and the motion detection program module, and the user is If the face moves too much, an alert is sent out via the alert unit.

1‧‧‧Image recognition based monitoring device

11‧‧‧Image capture unit

12‧‧‧ memory

13‧‧‧Processing unit

14‧‧‧Warning unit

15‧‧‧Pre-Processing Module

16‧‧‧Face Detection Program Module

17‧‧‧Motion detection program module

2‧‧‧Monitoring procedures

21~25‧‧‧Steps

Other features and effects of the present invention will be apparent from the following description of the drawings, wherein FIG. 1 is a block diagram illustrating a preferred embodiment of the image recognition based monitoring device of the present invention; and FIG. 2 is A flow chart illustrating a supervisor performed in the preferred embodiment Control program.

The preferred embodiment of the novel image recognition based monitoring device is suitable for monitoring a user of a bed, such as an infant or an elderly person who is bedridden. Referring to FIG. 1 , the monitoring device 1 includes an image capturing unit 11 , a memory 12 , a processing unit 13 , a warning unit 14 , a pre-processing program module 15 , a facial detection program module 16 , and a Motion detection program module 17. The memory 12 is a non-volatile memory such as a flash memory or an electronic erasable rewritable read-only memory, and is used to store the pre-processing program module 15 and the face detection program module 16 and The motion detection program module 17 is configured to execute the pre-processing program module 15, the face detection program module 16 and the motion detection program module 17; the warning unit is a speaker.

Referring to Figures 1 and 2, the monitoring device 1 is used to perform a monitoring program 2. The monitoring program 2 first performs the step 21, wherein the image capturing unit 11 captures a continuous first image and a second image of the user, wherein the capturing of the second image occurs in the first image. After picking up.

Then, in step 22, the pre-processing module 15 performs noise suppression on the first and second images. For example, the pre-processing module 15 uses a median filter to perform noise suppression.

Then, in step 23, the face detection module 16 performs face detection on the first and second images to detect a first facial block of the user in the first image and a second facial block of the user in the second image. The method of face detection can be utilized, for example, in an image. The face of the human body is defined by the division of the skin color.

Then, in step 24, the motion detection program module 17 performs facial motion detection on the first and second images to detect whether the second facial block is relative to the first facial block. The degree of movement of the face exceeds a preset threshold. The manner of detecting the motion of the face can be implemented, for example, by background subtraction, and the degree of movement of the face is calculated by comparing the difference of each pixel of the face block in the two images before and after, or by A background model is obtained by using the face block of the previous image, and then the face of the current image is compared with the background model to calculate the degree of movement of the face.

Then, if the detection result of the step 24 is negative, the step 21 is performed; and if the detection result of the step 24 is YES, the step 25 is performed, that is, the processing unit 13 triggers the warning unit 14 to send an outward one. Alert sounds. Although the alert unit 14 is a speaker in the preferred embodiment, it is obvious that the implementation of the alert unit 14 is not limited to the speaker.

The monitoring device 1 is particularly suitable for the safe monitoring of infants or elderly people who are bedridden. For example, when the baby turns to sleep from the lying state, the monitoring device 1 can detect that the baby's face is moved too much in the continuous image and send a warning to the outside, so that the caregiver can notice the state of the baby. Avoid sudden death.

In summary, the novel image recognition-based monitoring device 1 can detect the degree of facial movement of the user, and send an alarm to the outside through the warning unit 14 when the user's face moves excessively. News, Therefore, the purpose of this new type can be achieved.

However, the above is only a preferred embodiment of the present invention, and is not intended to limit the scope of the present invention, that is, the simple equivalent changes and modifications made in accordance with the scope of the present patent application and the contents of the patent specification, All remain within the scope of this new patent.

1‧‧‧Image recognition based monitoring device

11‧‧‧Image capture unit

12‧‧‧ memory

13‧‧‧Processing unit

14‧‧‧Warning unit

15‧‧‧Pre-Processing Module

16‧‧‧Face Detection Program Module

17‧‧‧Motion detection program module

Claims (5)

An image recognition-based monitoring device is suitable for monitoring a user of a bed, the device comprising: an image capturing unit for capturing a continuous first image and a second image of the user, wherein the The capture of the second image occurs after the capture of the first image; a memory for storing a face detection program module and a motion detection program module, wherein the face detection program module is used The motion detection program module is configured to detect whether a first face block of the user in the first image and a second face block of the user in the second image are detected. The movement of the second facial block relative to the face of the first facial block exceeds a predetermined threshold; a processing unit for executing the facial detection program module and the motion detection program And a warning unit, when the motion detection program module detects that the movement of the second facial block relative to the face of the first facial block exceeds the preset threshold, The processing unit triggers the alert unit to send an alert to the outside. The image recognition-based monitoring device of claim 1, further comprising a pre-processing program module stored in the memory for performing noise suppression on the first and second images, wherein the processing unit further uses To execute the pre-processing program module. The image recognition-based monitoring device of claim 1, wherein the face detection program module detects the human body in the first and second images The first and second face blocks are detected by the skin color area. The image recognition-based monitoring device of claim 1, wherein the motion detection program module uses a background subtraction method to detect whether the second facial block is opposite to the face of the first facial block. The degree of movement exceeds the preset threshold. The image recognition-based monitoring device of claim 1, wherein the alert unit is a speaker, and the alert is an alarm sound.