CN105138947A - Guard reminding method, reminding device and reminding system - Google Patents

Abstract

The invention provides a guard reminder method, a reminder device and a reminder system. After the image sequence of the person standing above the platform is acquired through the image acquisition unit, the image sequence is analyzed to determine whether the person is in a sleeping state. When a person is sleeping, a reminder signal is generated to alert the person.

Description

Guard reminding method, reminding device and reminding system

technical field

The present invention relates to a warning technology, and in particular to a guard reminding method for preventing people from drowsiness, a reminding device and a reminding system.

Background technique

On duty is one of the main tasks of the sentinel, and through the duty of the sentinel, emergencies can be discovered and dealt with in a timely manner. If there is any mistake on duty, it may have serious consequences for the current social security. Therefore, in order to ensure the safety of the society, the duty status of the sentries when they are on duty is very important.

At present, the management of the duty status of the sentries is mainly monitored by human means, such as regular inspections of the duty status of the sentries by superiors, or indirect monitoring methods, such as training or disciplinary constraints, to supervise the sentries to keep their posts. . However, the method cannot ensure that the sentinel is in a better state of duty, so often the sentinel still has a short sleep or trance.

Contents of the invention

The invention provides a guard reminding method, a guard reminding device and a guard reminding system, which can remind people standing on the platform to prevent them from falling asleep.

The present invention proposes a reminder method, which includes: acquiring an image sequence of a person standing above the platform through an image acquisition unit, wherein the image acquisition unit faces the platform and acquires the image sequence at a preset angle; analyzing the image sequence to determine whether the person presents sleep state; and when the person is in the sleep state, a reminder signal is generated to warn the person.

In one embodiment of the present invention, when the person is in a sleeping state, an early warning signal is sent to the remote monitoring host.

In one embodiment of the present invention, the step of judging whether the person is in a sleep state includes: detecting whether the person maintains a fixed posture for a preset time according to the image sequence, and if the person maintains a fixed posture for a preset time, then judge whether the person is in a sleep state. Show sleep state.

In an embodiment of the present invention, the step of judging whether the person is in a sleep state further includes: detecting facial objects in the image sequence, so as to judge whether the person's head does not turn within a preset time. Show sleep state.

In an embodiment of the present invention, when it is detected that the person's head is lowered downward, it is determined that the person is in a sleep state.

In an embodiment of the present invention, the step of judging whether the person is in a sleep state further includes: analyzing the image sequence to judge whether the person's eyes are in a closed state; and if the person's eyes are in a closed state, then judging that the person is in a sleep state .

In an embodiment of the present invention, the step of analyzing the image sequence to determine whether the person's eyes are closed eyes includes: executing a face recognition program on the image sequence to obtain facial objects in the image sequence; A search procedure is used to obtain the eye area in the face object; and it is judged whether the eyes are closed according to the size of the eye area.

In an embodiment of the present invention, after the step of generating the reminder signal, it further includes: analyzing the facial feature information of the image sequence to determine whether the person is gazing at the designated direction; and if the person is not gazing at the designated direction, continuously driving The warning unit sends out a warning signal until a person is detected looking at a designated direction.

In an embodiment of the present invention, the reminder signal includes vibration warning, light source warning or sound source warning.

The present invention further provides a guard reminding device, which includes a warning unit, an image acquisition unit, a storage unit and a processing unit. The warning unit is arranged in the platform. The image acquisition unit faces the platform and acquires an image sequence of a person standing on the platform at a preset angle. The storage unit has an image processing module. The processing unit is coupled between the warning unit, the storage unit and the image acquisition unit. The processing unit executes the image processing module to analyze the image sequence to detect whether the person is in a sleep state. If the person is in a sleeping state, the processing unit sends a control signal to the warning unit, so that the warning unit generates a reminder signal to warn the person.

The present invention also proposes a guard reminder system, which includes a platform, a warning unit, an image acquisition unit and a remote monitoring host. A platform for a person to stand on. The warning unit is arranged in the platform. The image acquisition unit faces the platform and acquires the image sequence of the person at a preset angle. The remote monitoring host is coupled between the warning unit and the image acquisition unit, receives and according to the image sequence from the image acquisition unit, to detect whether the person is in a sleep state. If the person is in a sleeping state, the remote monitoring host sends a control signal to the warning unit, so that the warning unit generates a reminder signal to warn the person.

Based on the above, if the posture of the person remains unchanged within the preset time, the warning unit in the platform will send a reminder signal to remind the person, so as to prevent the person from falling asleep.

In order to make the above-mentioned features and advantages of the present invention more comprehensible, the following specific embodiments are described in detail with reference to the accompanying drawings.

Description of drawings

Fig. 1 is a block diagram of a guard reminding device according to a first embodiment of the present invention;

Fig. 2 is a schematic diagram of installation of a guard reminding device according to the first embodiment of the present invention;

Fig. 3A is a method flowchart of a guard reminding method according to the first embodiment of the present invention;

3B is a schematic diagram of an image processing module according to the first embodiment of the present invention;

4 is a schematic diagram of an image processing module according to a second embodiment of the present invention;

Fig. 5 is a method flowchart of a guard reminding method according to a second embodiment of the present invention;

Fig. 6 is a schematic diagram of head rotation according to a second embodiment of the present invention;

7 is a schematic diagram of an image processing module according to a third embodiment of the present invention;

Fig. 8 is a method flowchart of a guard reminding method according to a third embodiment of the present invention;

Fig. 9 is a schematic diagram of an eye region according to a third embodiment of the present invention;

10 is a schematic diagram of an image processing module according to a fourth embodiment of the present invention;

Fig. 11 is a method flowchart of a guard reminding method according to a fourth embodiment of the present invention;

FIG. 12 is a block diagram of a guard reminder system according to a fifth embodiment of the present invention.

Explanation of reference signs:

10: personnel;

100: guard reminder device;

110, 1210: warning unit;

120, 1220: image acquisition unit;

1230: remote monitoring host;

130: storage unit;

132: database;

134: image processing module;

140: processing unit;

302: behavior detection module;

402, 702, 1002: face detection module;

404: facing the detection module;

610: head;

704: Closed eyes detection module;

902: eye area;

1004: sight detection module;

θ: preset angle;

d1: the first direction;

d2: second direction

I: image;

L: horizontal line;

N1: first central point;

N2: second central point;

P: platform;

R: reference point;

S302-S306, S502-S510, S802-S808, S1102-S1110: each step of the guard reminding method.

Detailed ways

For the sentry or duty personnel on guard, if the status of the sentry can be monitored in real time, and a reminder is issued when the sentry or duty personnel are drowsy or in a daze, it is bound to ensure that the sentry or duty personnel are on duty. The present invention is a guard reminder method, a guard reminder device, a guard reminder system, and a computer program product based on the above viewpoints. In order to make the content of the present invention clearer, the following specific examples are given as examples in which the present invention can actually be implemented.

first embodiment

FIG. 1 is a block diagram of a guard reminding device according to a first embodiment of the present invention. Referring to FIG. 1 , the guard reminder device 100 includes a warning unit 110 , an image acquisition unit 120 , a storage unit 130 and a processing unit 140 . The processing unit 140 is coupled to the warning unit 110 , the image acquisition unit 120 and the storage unit 130 . The processing unit 140 utilizes the image acquisition unit 120 to monitor the waking state of the person, so as to determine whether to drive the alarm unit 110 to send a reminder signal to remind the person, so as to prevent the person from falling asleep. In this embodiment, the warning unit 110 , the storage unit 130 and the processing unit 140 are arranged in a platform, and the image acquisition unit 120 is arranged around the platform. An example is given below to illustrate the installation of the guard reminding device 100 .

Fig. 2 is a schematic diagram of installation of the guard reminding device according to the first embodiment of the present invention. Please refer to FIG. 2, the warning unit 110, the storage unit 130 and the processing unit 140 are arranged in the platform P, wherein the lens of the image acquisition unit 120 faces the platform P and acquires an image sequence at a preset angle, for example, the image acquisition unit 120 will It is set facing upwards with a preset angle θ between it and the horizontal line L, so as to acquire an image sequence of the person 10 standing on the platform P. The data transmission between the processing unit 140 and the image acquisition unit 120 can transmit the image sequence through wireless transmission or wired transmission.

The warning unit 110 includes but not limited to vibration warning, light source warning or sound source warning devices, and the warning signal it sends out is, for example, light source warning, vibration warning or sound source warning, so as to remind the personnel 10 on the platform P. However, the present embodiment is not limited thereto, and any device capable of sending out a warning signal can be used as the warning unit 110 .

The image acquisition unit 120 is used to acquire an image sequence of the person 10 , that is, to acquire multiple images continuously. The image acquisition unit 120 is, for example, a video camera or a camera using a Charge Coupled Device (CCD) lens, a Complementary Metal Oxide Semiconductor Transistor (CMOS) lens, or an infrared lens, but is not limited thereto.

The storage unit 130 includes a database 132 and an image processing module 134 . The database 132 is used to store the image sequences obtained by the image acquisition unit 120 . The image processing module 134 is, for example, an application program. The image processing module 134 is executed by the processing unit 140 to realize the analysis of the posture or the change of the local area of the person 10 . The storage unit 130 is, for example, a non-volatile memory (Non-volatile memory), a random access memory (Random Access Memory, RAM), or a hard disk. The above are examples only and not limited thereto.

The processing unit 140 is coupled to the warning unit 110 , the image acquisition unit 120 and the storage unit 130 , and controls the overall operation of the guard reminding device 100 . Specifically, the processing unit 140 will execute the image processing module 134 to analyze the image sequence, and determine whether the person 10 maintains the same posture within a preset time, so as to determine whether to drive the alarm unit 110 to issue a reminder signal. The processing unit 140 includes a central processing unit (Central Processing Unit, CPU for short) or a microprocessor (microprocessor), but is not limited thereto.

Here, the present embodiment is realized by using program codes. That is, the image processing module 134 in the storage unit 130 may be composed of one or more program code segments, and these program code segments include a plurality of instructions, and the processing unit 140 will execute these instructions through the image processing module 134 to realize the guard Reminds multiple steps of the method.

In the following, each step of the guard reminding method will be described in detail in conjunction with the above-mentioned guard reminding device 100 . FIG. 3A is a flow chart of a guard reminding method according to the first embodiment of the present invention. Please refer to FIG. 1 to FIG. 3A at the same time. In step S302 , the image acquisition unit 120 acquires an image sequence of the person 10 standing on the platform P. Referring to FIG. For example, the preset angle θ between the image acquisition unit 120 and the horizontal line L of the ground is 45 degrees, but not limited thereto, and the preset angle θ can be adjusted according to the height of the person.

In step S304 , the processing unit 140 executes the image processing module 134 to analyze the image sequence to determine whether the person 10 is in a sleep state. Since the person 10 generally maintains the same posture when dozing off, the processing unit 140 can determine whether the person 10 is sleeping according to whether the person 10 maintains the same posture or not. For example, the image processing module 134 can detect whether the person 10 maintains a fixed posture for a preset time to determine whether the person 10 is in a sleep state.

For example, FIG. 3B is a schematic diagram of an image processing module according to the first embodiment of the present invention. As shown in FIG. 3B , the image processing module 134 may include a behavior detection module 302 for judging whether the person 10 maintains the behavior within a preset time. Fixed posture. Specifically, the behavior detection module 302 can obtain a plurality of feature points of the relevant person from the image sequence, and judge whether the person 10 maintains a fixed posture according to changes of these feature points within a preset time. For example, the behavior detection module 302 can compare the distribution difference of feature points in two successive images within a preset time, and determine whether the distribution difference is smaller than a preset value. If the distribution difference is smaller than the preset value, that is, the images of the person 10 in the two successive images are similar, the behavior detection module 302 will determine that the person 10 maintains a fixed posture for a preset time. If the distribution difference of the feature points in the two successive images is not less than the preset value, that is, the images of the person 10 in the two successive images are not similar, the behavior detection module 302 will determine that the person 10 has moved within the preset time. Accordingly, if the person 10 maintains a fixed posture for a preset time, the processing unit 140 can determine that the person 10 is in a sleep state according to the behavior detection module 302 . Conversely, if the person 10 moves within the preset time, for example, the head, limbs, eyes or other areas of the person 10 do not maintain a fixed posture, it means that the person is not dozing off, so the processing unit 140 can The detection module 302 determines that the person 10 is not in a sleep state.

Returning to FIG. 3A , if the person 10 is in a sleeping state, as shown in step S306 , the processing unit 140 will drive the warning unit 110 to send out a reminder signal to warn the person 10 . Alternatively, the processing unit 140 can also send an early warning signal to a remote monitoring host, so as to alert the personnel 10 by sending a reminder signal from the remote monitoring host. In this way, when the person 10 is dozing off, the warning unit 110 in the platform P will send a reminder signal such as a vibration signal or a sound signal to remind the person 10, so that the person 10 can be kept awake.

If the person 10 is not in a sleep state, the image acquisition unit 120 will continue to acquire the image sequence of the person 10 standing above the platform P (step S302), that is, the guard reminding device 100 will not drive the alarm unit 110 to remind the person 10 at this time. .

The processing unit 140 executes the image processing module 134 to analyze the image sequence to determine whether the person 10 maintains the same posture within a preset time, and the processing unit 140 also judges according to whether the head of the person 10 is turning or not. Because when the person 10 is awake, his head will swing from side to side to check the surrounding environment, so whether the person 10 maintains the same posture can be judged by whether the head turns or not, and whether the person 10 is dozing off can be quickly detected. Another example will be given below for illustration.

second embodiment

FIG. 4 is a schematic diagram of an image processing module according to a second embodiment of the present invention. Please refer to FIG. 4 , the image processing module 134 further includes a face detection module 402 and an orientation detection module 404 . The face detection module 402 is used to identify facial objects in the image sequence, and the orientation detection module 404 detects the movement state of the head according to the changes of the facial objects.

Fig. 5 is a flow chart of a guard reminding method according to the second embodiment of the present invention. The following description will be made in conjunction with the guard reminding device 100 in FIG. 1 . Please refer to FIG. 1 , FIG. 4 and FIG. 5 at the same time. In step S502 , the image acquisition unit 120 acquires an image sequence of the person 10 standing on the platform P. Referring to FIG.

After the processing unit 140 receives the image sequence from the image acquisition unit 120 , in step S504 , the processing unit 140 executes the image processing module 134 to analyze the image sequence. Among them, the face detection module 402 will execute the face recognition program on the image sequence to obtain facial objects in the image sequence, and analyze the changes of the facial objects, and the orientation detection module 404 will be based on the facial object in the image sequence. change, and detect whether the head of the person 10 turns.

In detail, the face detection module 402 can use the AdaBoost algorithm or other existing face detection algorithms (eg, use Haar-like features to perform face recognition actions) to obtain the facial objects in each image. In addition, the face detection module 402 can also search for the nostril area in the face object, that is, the positions of the two nostrils, and obtain the nostril position information, such as the first center point and the second center point of the two nostrils. Afterwards, the facing detection module 404 may determine whether the head turns based on the nostril position information.

For example, FIG. 6 is a schematic diagram of head rotation according to the second embodiment of the present invention. Please refer to FIG. 4 and FIG. 6. After the face detection module 402 recognizes the facial object (as shown in image I), the facing detection module 404 can use the first center point N1 and the second center point N2 of the two nostrils to determine Whether the head 610 in the image turns to the first direction d1 or turns to the second direction d2. Here, in terms of the direction of the person 10, the right direction is regarded as the first direction d1, and the left direction is regarded as the second direction d2. Specifically, the facing detection module 404 can compare the first center point N1 and the second center point N2 with the reference point R, and through the positional relationship between the first center point N1 and the reference point R and the second center point N2 Based on the positional relationship with the reference point R, it is determined in which direction the head 610 turns. Of course, the facing detection module 404 can also use the first center point N1 or the second center point N2 to perform head-down detection. For example, the facing detection module 404 can compare one of the first central point N1 or the second central point N2 with the reference point R to determine whether the first central point N1 or the second central point N2 is located below the reference point R , and when the first central point N1 or the second central point N2 is located below the reference point R, the facing detection module 404 can determine that the head 610 is lowered.

In this way, the processing unit 140 can determine whether the person 10 maintains the same posture within a preset time according to the detection result of the facing detection module 404 .

Returning to FIG. 5, when the facing detection module 404 detects that the head of the person 10 has not turned within the preset time, as shown in step S506, the processing unit 140 will determine that the person 10 has not maintained the same head within the preset time. posture, and in step S508, the processing unit 140 will drive the alarm unit 110 to send out a reminder signal to remind personnel. In other words, since the head of the person 10 does not turn when he is dozing off, when the facing detection module 404 does not detect the head turning of the person 10 within a preset time, the processing unit 140 will regard the person 10 as dozing off , and then keep the personnel awake through the warning unit 110 .

If the face detection module 404 detects that the head of the person 10 has turned within the preset time, as shown in step S510, the processing unit 140 will determine that the person does not maintain the same posture within the preset time, and in step S502, the image The acquisition unit 120 will continue to acquire the sequence of images of the person 10 standing on the platform P. That is, at this time, the processing unit 140 will regard the person 10 as awake, so the person 10 will not be reminded by the warning unit 110 , and the person 10 will continue to be monitored by the image acquisition unit 120 .

As mentioned above, if the person 10 maintains the same posture within the preset time, the processing unit 140 will also judge more accurately whether the person 10 is dozing off according to whether the person 10's eyes are opened or closed, and then decide whether to drive the warning unit 110 to send a reminder signal . Another example is given below for illustration.

third embodiment

FIG. 7 is a schematic diagram of an image processing module according to a third embodiment of the present invention. Please refer to FIG. 7 , the image processing module 134 includes a face detection module 702 and a closed-eye detection module 704 . Wherein, the face detection module 702 is similar to the face detection module 402 of the second embodiment, so details are omitted here. The eye-closed detection module 704 is used for executing an eye search program on the face object to determine whether the eyes are closed.

Fig. 8 is a method flowchart of a guard reminding method according to a third embodiment of the present invention. The following will be described in conjunction with the guard reminding device 100 in FIG. 1 . Please refer to FIG. 1 , FIG. 7 and FIG. 8 at the same time. In step S802 , the image acquisition unit 120 acquires an image sequence of the person 10 standing on the platform P. Referring to FIG. In step S804 , the processing unit 140 executes the image processing module 134 to analyze the image sequence to determine whether the person 10 maintains the same posture within a preset time.

Different from the foregoing embodiments, if the person 10 maintains the same posture within a preset time, the image processing module 134 will further determine whether the eyes of the person 10 are closed (step S806 ). For example, the face detection module 702 will perform a face recognition program on the image sequence to obtain facial objects in the image sequence, and the closed-eye detection module 704 will perform an eye search program on the facial objects to obtain facial objects in the image sequence. Obtain the eye area from the object, and then judge whether the eyes are closed according to the size of the eye area.

In detail, after the face detection module 702 obtains the face object, it can further search the nostril area in the face object to obtain the nostril position information. Next, the closed-eye detection module 704 can estimate the eye search frame according to the nostril position information, so as to detect the eye region within the eye search frame. That is to say, the nostrils are easier to identify in the image than the eyes. Therefore, after the nostrils are found first, the eye search frame is estimated upwards to find the eye area in the eye search frame. This narrows the eye search range.

Furthermore, after the eye-closed detection module 704 executes the eye search program to extract the eye region, the contrast of the eye region is adjusted to obtain an enhanced image. Next, denoising, edge sharpening, binarization, and edge sharpening are sequentially performed on the enhanced image to obtain an image of the eye region 902 as shown in FIG. 9 . Fig. 9 is a schematic diagram of an eye region according to a third embodiment of the present invention.

Next, the closed eye detection module 704 can determine whether the eyes are closed based on the size of the eye area 902 , thereby obtaining eye opening and closing information. For example, when the height of the eye region 902 is smaller than the height threshold (for example, the range of the height threshold is within 5 to 7 pixels), and the width of the eye region 902 is greater than the width threshold (for example, the range of the width threshold is within At 60-80 pixels), it is determined that the eyes are currently closed. If the above conditions are not met, it is determined that the eyes are currently in an open state.

Returning to FIG. 8 , when the eyes of the person 10 are closed, as shown in step S808 , the processing unit 140 will drive the warning unit 110 to send out a warning signal. In this embodiment, the closed-eye detection module 704 will continue to judge whether the eyes of the person 10 are still in the closed-eye state (as shown in step S806), and before the closed-eye detection module 704 detects that the personnel's eyes are in the open-eye state , the processing unit 140 will continue to drive the warning unit 110 to send out a reminder signal. In other words, until the eye-closed detection module 704 detects that the eyes of the person 10 are in a non-closed state, that is, the eyes of the person 10 are in an open state, the processing unit 140 will stop driving the warning unit 110 to send a reminder signal, and in step S802, the image acquisition unit 120 continues to acquire the image sequence of the person 10 .

It is worth mentioning that the reminder signal sent by the warning unit 110 is a signal for notifying the personnel 10 to watch a designated direction. For example, a voice signal is used to notify the person to watch the image acquisition unit 120 to ensure that the eyes of the person 10 are not closed. Of course, a vibration signal can also be used to remind personnel to look in a designated direction. Another example is given below for illustration.

Fourth embodiment

FIG. 10 is a schematic diagram of an image processing module according to a fourth embodiment of the present invention. Referring to FIG. 10 , the image processing module 134 includes a face detection module 1002 and a line of sight detection module 1004 . Wherein, the face detection module 1002 is similar to the face detection module 402 of the second embodiment, so details will not be repeated here. The line of sight detection module 1004 is used to analyze the facial feature information of the facial object to determine the gaze direction of the person.

Fig. 11 is a flow chart of a guard reminding method according to a fourth embodiment of the present invention. The following will be described in conjunction with the guard reminding device 100 in FIG. 1 . Please refer to FIG. 1 , FIG. 10 and FIG. 11 at the same time. In step S1102 , the image acquisition unit 120 acquires an image sequence of the person 10 standing on the platform P. Referring to FIG. In step S1104 , the processing unit 140 executes the image processing module 134 to analyze the image sequence to determine whether the person 10 maintains the same posture within a preset time. If the person 10 changes posture, return to step S1102, and the image acquisition unit 120 continues to acquire the image sequence of the person 10. On the contrary, if the person 10 maintains the same posture within the preset time, as shown in step S1106 , the processing unit 140 will drive the alarm unit 110 to send out a reminder signal.

Different from the foregoing embodiments, the warning signal sent by the warning unit 110 is used to require the person 10 to look at a designated direction. For example, a voice signal is sent out through the warning unit 110 to notify the person 10 to look at the image acquisition unit 120 or other designated directions such as upwards, downwards, etc. Alternatively, the warning unit 110 sends out a vibration signal, so that the person 10 realizes that the vibration signal should be released by gazing at a designated direction.

Next, in step S1108 , the image processing module 134 analyzes the facial feature information of the image sequence to determine whether the person 10 is looking at a specified direction. For example, the facial object in the image sequence is found by the face detection module 1002 , so that the gaze detection module 1004 can analyze the facial feature information of the facial object to determine the gaze direction of the person 10 . For example, the line of sight detection module 1004 finds the eye area from the facial object, and then tracks the movement of the pupil position to calculate the eyeball movement coordinates. In this way, the gaze detection module 1004 can obtain the current gaze direction of the person 10 according to the eyeball movement coordinates.

If the person 10 is not looking at the designated direction, as shown in step S1106 , the processing unit 140 will continue to drive the warning unit 110 to send out a warning signal. If the line of sight detection module 1004 detects that the person 10 is gazing at a specified direction, as shown in step S1110 , the processing unit 140 will make the warning unit 110 stop sending out the warning signal. In other words, when the person 10 is not looking at the specified direction, the processing unit 140 will continue to drive the warning unit 110 to send out a reminder signal until the line of sight detection module 1004 detects that the person is looking at the specified direction, and the processing unit 140 will stop the warning unit 110 Send a reminder signal. In this way, in order to prevent the person 10 from still dozing off after the alarm unit 110 sends out a reminder signal, by judging whether the person 10 is looking at a designated direction, it can be more ensured that the person 10 returns to the awake state.

fifth embodiment

The above-mentioned guard reminding method can also be realized by a guard reminding system. FIG. 12 is a block diagram of a guard reminder system according to a fifth embodiment of the present invention. Please refer to FIG. 12 , the guard reminder system includes a platform P, a warning unit 1210 , an image acquisition unit 1220 and a remote monitoring host 1230 , wherein the remote monitoring host 1230 is coupled between the warning unit 1210 and the image acquisition unit 1220 . Here, the functions of the platform P, the warning unit 1210 and the image acquisition unit 1220 are the same or similar to those of the platform P, the warning unit 110 and the image acquisition unit 120 in the first embodiment, so details are not repeated here.

Different from the foregoing embodiments, this embodiment can use the remote monitoring host 1230 to implement the guard reminder method, wherein the remote monitoring host 1230 has a storage unit and a processing unit, wherein the storage unit has an image processing module, and the processing unit will pass The image processing module is used to execute each step of the guard reminding method. Here, the functions of the storage unit and the processing unit are the same as or similar to those of the above-mentioned embodiments.

In this way, the remote monitoring host 1230 receives and according to the image sequence from the image acquisition unit 1220 , detects whether the person 10 is in a sleeping state. Moreover, if the person 10 is in a sleeping state, the remote monitoring host 1230 will send a control signal to the warning unit 1210 to make the warning unit 1210 generate a reminder signal to warn the person 10 . In addition, the remote monitoring host 1230 can also judge whether the person 10 maintains a fixed posture and whether the person 10 is closed or not according to the internally stored behavior detection module, face detection module, orientation detection module, line of sight detection module and eye-closed detection module. Eye state, judging whether the head of the person 10 turns or lowers his head, and judging whether the person 10 is staring at the designated direction, so as to judge whether the person 10 is dozing off, and generate a reminder signal to keep the person 10 awake, and the detailed steps of the guard reminder method Reference can be made to the above-mentioned embodiments, so details are not repeated here.

Sixth embodiment

The present invention further provides a computer program product, which includes a computer-readable recording medium. A computer-readable recording medium records several program instructions. After these program instructions are loaded into the electronic device and executed, the steps of the above-mentioned guard reminder method can be completed, and at the same time, the electronic device can realize the functions described in the above-mentioned embodiments. Wherein, the computer-readable recording medium may be a read-only memory, a random access memory, a magnetic tape, a floppy disk, a hard disk, an optical disc, transmission media, and the like. In other embodiments, the computer program product may also be provided directly via a network, and the scope thereof is not limited here.

To sum up, in the guard reminding method, guard reminding device, guard reminding system and computer program product, the guard reminding device will detect whether the person's posture changes within a preset time. If the posture of the person remains unchanged within the preset time, the warning unit in the platform will send out a reminder signal to remind the person, so as to prevent the person from falling asleep. In addition, the alarm unit can continue to send out a reminder signal until the person's posture changes or the person looks at the specified direction, thereby ensuring that the person returns to the awake state. In this way, the guard reminder device can monitor the status of the person standing on the platform in real time, and when the person is dozing off or in a daze, the guard reminder device can quickly remind the person, so as to prevent the person from falling asleep and keep the person awake .

Finally, it should be noted that: the above embodiments are only used to illustrate the technical solutions of the present invention, rather than limiting them; although the present invention has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that: It is still possible to modify the technical solutions described in the foregoing embodiments, or perform equivalent replacements for some or all of the technical features; and these modifications or replacements do not make the essence of the corresponding technical solutions deviate from the technical solutions of the various embodiments of the present invention. scope.

Claims (19)

1. A guard reminding method is characterized in that, comprising: acquiring an image sequence of a person standing above a platform by an image acquisition unit, wherein the image acquisition unit faces the platform and acquires the image sequence at a preset angle; analyzing the image sequence to determine whether the person exhibits a sleep state; and When the person is in the sleeping state, a reminder signal is generated to warn the person. 2. The guard reminding method according to claim 1, wherein when the person is in the sleep state, an early warning signal is sent to a remote monitoring host. 3. The guard reminding method according to claim 1, wherein the step of judging whether the person presents the sleep state comprises: According to the image sequence, it is detected whether the person maintains a fixed posture within a preset time, and if the person maintains the fixed posture within the preset time, it is determined that the person is in the sleep state. 4. The guard reminding method according to claim 1, wherein the step of judging whether the person presents the sleep state comprises: Detecting a facial object of the image sequence to determine that the person is in the sleep state if a head of the person is detected not to be turned within the preset time; and When it is detected that the head of the person is lowered downward, it is determined that the person is in the sleep state. 5. The guard reminding method according to claim 1, wherein the step of judging whether the person presents the sleep state also includes: executing a face recognition program on the image sequence to obtain a facial object in the image sequence; Execute an eye search procedure to obtain an eye area in the face object; judging whether the eye is in the closed eye state based on the size of the eye area; and If the eyes of the person are in the eye-closed state, it is judged that the person is in a sleep state. 6. The guard reminder method according to claim 1, characterized in that, after the step of generating the reminder signal, comprising: Analyzing a facial feature information of the image sequence to determine whether the person is looking at a specified direction; and If the person is not gazing at the designated direction, the warning unit is continuously driven to send out the reminder signal until it is detected that the person is gazing at the designated direction. 7. The guard reminder method according to claim 1, wherein the reminder signal includes a vibration warning, a light source warning or a sound source warning. 8. A guard reminding device, characterized in that it comprises: a warning unit, set in a platform; an image acquisition unit facing the platform and acquiring an image sequence of a person standing above the platform at a preset angle; a storage unit having an image processing module; and A processing unit, coupled between the warning unit, the storage unit and the image acquisition unit, the processing unit executes the image processing module to analyze the image sequence to detect whether the person is in a sleep state; If the person is in the sleep state, the processing unit sends a control signal to the warning unit, so that the warning unit generates a reminder signal to warn the person. 9. The guard reminding device according to claim 8, wherein if the person is in the sleep state, the processing unit sends an early warning signal to a remote monitoring host. 10. The guard reminding device according to claim 8, wherein the image processing module comprises: A behavior detection module detects whether the person maintains a fixed posture within a preset time according to the image sequence. If the person maintains the fixed posture within the preset time, it is judged that the person is in the sleep state. 11. The guard reminding device according to claim 8, wherein the image processing module comprises: A human face detection module, which detects a facial object in the image sequence; A face detection module, according to the facial object in the image sequence, detects whether the head of the person has turned, and when it is detected that the head of the person does not turn within the preset time, the The processing unit determines that the person is in the sleeping state, wherein, when the facing detection module detects that the head of the person is lowered downward, the processing unit determines that the person is in the sleeping state; and a line-of-sight detection module, which analyzes a facial feature information of the facial object to determine whether the person is gazing in a specified direction, and to determine whether the person is in the sleeping state if the person is not gazing in the specified direction, And make the warning unit continue to send out the warning signal until it is detected that the person is staring at the designated direction. 12. The guard reminding device according to claim 8, wherein the image processing module comprises: A human face detection module, which detects a facial object in the image sequence; A closed-eye detection module, which executes an eye search program on the face object to obtain an eye area in the face object, and then judges whether the eye is in the closed-eye state according to the size of the eye area. When the eyes are closed, it is determined that the person is in a sleep state. 13. The guard reminder device according to claim 8, wherein the reminder signal includes a vibration warning, a light source warning or a sound source warning. 14. A guard reminder system, characterized in that it comprises: a platform on which a person stands; a warning unit, set in the platform; an image acquisition unit facing the platform and acquiring an image sequence of the person at a preset angle; and A remote monitoring host, coupled between the warning unit and the image acquisition unit, receives and according to the image sequence from the image acquisition unit, to detect whether the person is in a sleep state; Wherein, if the person is in the sleep state, the remote monitoring host sends a control signal to the warning unit, so that the warning unit generates a reminder signal to warn the person. 15. The guard reminder system according to claim 14, wherein the remote monitoring host includes: a storage unit having an image processing module; and A processing unit, coupled to the storage unit, the processing unit executes the image processing module to analyze the image sequence to determine whether the person is in the sleep state, and if the person is in the sleep state, then transmit the control signal to the alarm unit. 16. The guard reminder system according to claim 15, wherein the image processing module comprises: A behavior detection module detects whether the person maintains a fixed posture within a preset time according to the image sequence. If the person maintains the fixed posture within the preset time, it is judged that the person is in the sleep state. 17. The guard reminding system according to claim 16, wherein the image processing module comprises: A human face detection module, which detects a facial object in the image sequence; A face detection module detects whether the head of the person has turned according to the facial object in the image sequence, and if the head of the person does not turn within a preset time, the processing unit judges whether the head of the person has turned The person is in the sleeping state, wherein, when the facing detection module detects that the head of the person is lowered downward, the processing unit determines that the person is in the sleeping state; and a line-of-sight detection module, analyzing a facial feature information of the facial object to determine whether the person is looking at a specified direction; if the person is not looking at the specified direction, it is determined that the person is in the sleeping state, and The warning unit is made to continuously send out the warning signal until it is detected that the person gazes at the designated direction. 18. The guard reminding system according to claim 16, wherein the image processing module comprises: A human face detection module, which detects a facial object in the image sequence; A closed-eye detection module, which executes an eye search program on the face object to obtain an eye area in the face object, and then judges whether the eye is in the closed-eye state according to the size of the eye area. When the eyes are closed, it is determined that the person is in a sleep state. 19. The guard reminder system according to claim 14, wherein the warning signal includes a vibration warning, a light source warning or a sound source warning.