FI121198B - Method, system and computer software for monitoring a person - Google Patents

Description

Method, system and computer software for monitoring a person

Area

The invention relates to a method and system for monitoring 5 persons, and is therefore suitable, for example, for home care of a patient or other individual in home care.

Background

It is predicted that the average age will rise sharply in the coming years and that the proportion of the population over 65 will increase. This poses significant 10 challenges for the state, municipalities and private bodies in relation to elderly care. The goal would be for as many older people as possible to live independently in their own home and familiar environment for as long as possible. This would be both cost effective and meaningful for the elderly themselves.

Prior art provides a variety of methods for monitoring patient movements and functions that allow the home or elderly person to be treated at home. Commercial use includes e.g. Devices attached to the patient's body (e.g., wrist) that monitor the patient's vital signs, movement, activity, etc., and alert when the patient's abnormalities are detected. The problem with these patient-attached devices is that they interfere with the patient's normal life (including washing), and the patient may disconnect such device. This also undermines the reliability of patient monitoring. These devices are usually based on the patient's own alert, which is unreliable because the patient himself may not be the right person to decide on the need for the alert, and the patient may not be able or willing to make the alert, even if necessary.

Motion sensors or video cameras placed in the patient's home or similar space are also shown to monitor the patient's movement in the monitored state. The problem with motion sensors is that they are only able to tell where the patient is moving in the part of the monitored space and do not give a reliable picture of the patient's positions or movements. They are not capable of, for example. to tell if the patient has fallen. Camcorders can detect such information when processing a video image with suitable video processing equipment, but it is difficult to identify a person from a video image because the human body 2 may be confused with the background. In addition, the video image is very sensitive to light and it is not possible to observe a person's movements at night, for example.

Short description

It is an object of the invention to provide a method and a system implementing method-5 such that patient monitoring can be performed reliably.

The invention relates to a process according to claim 1.

The invention further relates to a system according to claim 6, adapted to carry out the method according to claim 1.

The invention also relates to a software product according to claim 11, which is adapted to be executed in the system according to claim 6.

Preferred embodiments of the invention are described in the dependent claims.

Several advantages are achieved by the method and system of the invention. The invention utilizes thermal cameras for patient monitoring. The image quality of the thermal imaging camera is completely independent of the illumination and can be easily detected by the image due to the good temperature resolution of the thermal imaging image and the passive imaging technology. This also makes the person clearly distinguishable from the background, which facilitates the recognition of movements and postures and reduces the need for image processing power. Also, the patient does not need to carry any special device, which improves the reliability of the system. The system is capable of independently triggering an alarm based on a detected event, and alarms can be categorized based on detected ta-25 events. Thermal imaging surveillance can also be combined with, for example, fire and burglar alarms, stove watchers, etc. watches, which improve home security. In addition, the system is easily and inexpensively portable and configurable to any location (apartment or facility). The invention also enables statistical monitoring of patient activity and lifestyle, which facilitates patient care.

List of figures

The invention will now be described in more detail in connection with preferred embodiments, with reference to the accompanying drawings, in which: Figure 1 shows a system according to an embodiment of the invention when installed in a dwelling; Fig. 2 shows a block diagram of Figure 5 shows an analysis of a thermal camera image filtered in accordance with one embodiment of the invention, Figure 5 shows an event related to a thermal camera image analyzed in accordance with one embodiment of the invention, Figures 6A and 6B show a flow diagram of illustrates a process for performing patient monitoring in accordance with an embodiment of the invention, and FIG. 8 is a flow chart illustrating the invention; a process according to one embodiment for detecting a patient falling over.

15 Description of Embodiments

The patient monitoring system of the present invention includes one or more thermal cameras, the image of which can be used to accurately detect a person, regardless of lighting conditions. As mentioned above, conventional camcorders do not work for this purpose in the dark. Night vision cameras based on infrared light have the same problem, even if they operate in the dark. The problem with an infrared light source based night vision camera is that not only the human body but also the environment reflects the radiation produced by the light source, and thus distinguishing a person from the image recorded by the night vision device can be very difficult, which undermines the reliability of such a system.

The thermal camera, on the other hand, is passive and relies on the recording of thermal radiation emitted by the subject itself, and thus has a significantly better resolution than an IR-based device, since the resolution is not impaired by the infrared light reflected by the environment. 30 In a typical indoor environment, the environment reflects heat radiation corresponding to room air temperature, so that a 37-degree (celsius) human body can be easily distinguished from its surroundings by a clearly warm object. Domestic animals and various electrical appliances also emit radiation warmer than indoor air, but the human body can be distinguished from these by signal processing based on the shape of the human body. The separation of the human body from the image by machine vision based signal processing is known per se in the art. Also, the recognition of the positions of the human 4 observed in the image is known per se in the field of technology related to machine vision. Typically, a thermal imaging image is processed by a signal processing apparatus by searching for a human image in the stored thermal imaging image and comparing the parameters describing the positions of the human image found in the image with the stored reference parameters of corresponding positions. The human figure refers to the person detected in the thermal camera image. Thus, a thermal camera image identifies a human figure, but no attempt is made to identify the human identity in order to maintain privacy. Depending on the embodiment, the thermal camera image produced by the thermal camera may be a still image, a series of still images, or a video image. 10 The implementation depends on how closely the subject's movements are to be monitored, the capacity of the system's signal processing equipment, etc.

Figure 1 illustrates a system according to an embodiment of the invention installed in an apartment for patient monitoring. For example, the apartment may be the patient's home, whereby the system allows for the patient's home care, 15 which is less expensive than in-patient care and meaningful to the patient. The patient may be, for example, an elderly person or a person suffering from dementia or other mental illness or physical disability who needs regular treatment. The invention can also be applied to private use, for example to improve home security.

20 In this system, a number of thermal cameras 100 to 108 are installed in the apartment so that they can observe most of the area of the apartment and especially the desired rooms or objects. For example, the thermal camera 106 is directed to produce a thermal camera image of the kitchen, and in particular of the stove 110. This is capable of detecting a forgotten 25 plate of the stove and, if necessary, triggering an alarm, as described below. Thermal cameras do not necessarily have to cover the entire apartment. For example, you may not want to place a thermal imaging camera in the bathroom for the sake of maintaining intimacy protection, even though the thermal imaging never seeks to identify a person's identity. In addition, the apartment is provided with 30 signal processing equipment which analyzes the image produced by the thermal cameras by searching for a human figure in the thermal camera image and analyzing the positions and / or movement of the human image found in the image. The signal processing equipment may be located in the same room, apartment or building where the patient is to be monitored. Alternatively, the thermal camera image 35 may be transmitted over a communication network to a remote point where the thermal camera image analysis is performed. For example, the remote point may be located in a different building 5 than the observable space. The system of this embodiment of the invention also includes communication means for making a possible alert if an event is detected in the image analysis that requires a nurse to be alerted to attend to a patient (or other type of alert).

Figure 2 shows a block diagram of a patient monitoring system using thermal cameras, which may be arranged in the apartment as shown in Figure 1. As mentioned above, the system comprises one or more thermal cameras 100-110, signal processing equipment 202, and communication means 200. The communication means may include the necessary hardware and Ohio software to establish communication with the patient care and monitoring service provider. The communication connection can be established through a LAN connection to the LAN of the building, an xDSL connection (Digital Subscriber Line) or wirelessly utilizing, for example, a cellular network (GSM, UMTS, WiMAX). The signal processing apparatus 202 and the thermal cameras 15 100-110 may communicate with each other over a LAN connection (e.g., Ethernet).

The signal processing apparatus 202 is included in the system control unit 204 and is configured to receive from the thermal cameras 100 to 110 a snapshot of the observed state and analyze the snapshot to search for a human and to compare the parameters for interpreting movements and postures. Parameters corresponding to reference positions may be stored in a memory unit 206 which may include parameters for standing, sitting, falling, various cues (e.g., 25 waving hands), etc. If the signal processing unit 202 detects a match in the image recorded by a message to the control unit 204 to detect that position. The control unit 204 then performs a predetermined operation associated with that event 30 as a result of an event detected in the analysis, which may be an alarm and / or storing the event in the memory unit. The control unit 204 may be configured to perform a different operation depending on the detected event. For example, if a patient is found to have fallen but rises, the control unit can store the fall detection and event time 35 (date and time) in the memory unit. In this case, the control unit may be configured to report a crash through communication means 6 to a monitoring center connected to the system, so that the next routine visit by the caregiver can check if the patient is injured. In other words, the control unit 204 may perform a low priority alarm that does not require immediate action. On the other hand, if it is detected that the patient is lying on the floor 5 and is not moving (for example, due to overturning), the control unit 204 may be configured to perform a high priority alarm requiring immediate intervention by a caregiver or the like. The operation performed by the control unit 204 for each event detected by the analysis can be parameterized as desired. The control unit 104 may include in each alarm a message indicating a detected event in text or image format. The control unit 204 may be configured to store all detected events in the memory unit 206 along with the event time, which enables statistical monitoring of patient activity to be accurately performed over a long period of time. For example, signal processing equipment 202 may be arranged to report an detected event if a patient is detected walking in the middle of the night (or at a different time of day). In this way, the patient's sleep intake can be monitored.

The system of Figure 2 can also be implemented with one or more so-called smart cameras. A smart camera is called hardware that includes 20 cameras (in this case, a thermal camera) and a processing unit that can be configured by software to process the image produced by the camera. Referring to Figure 2, a system according to an embodiment of the invention may be implemented with one or more smart cameras, each smart camera comprising a thermal camera, signal processing equipment 202, control unit 204, 25 memory unit 206 and communication means 200 in one assembly (e.g., one housing). For example, the memory unit 206 may be a memory card removably connected to the smart camera and adapted to store software and parameters configuring the processing unit.

Figures 3A-3C illustrate a thermal camera image analyzed in accordance with an embodiment of the invention, showing the event of a patient falling over. The picture includes a human figure 302 corresponding to the patient and a background lamp 304. The lamp 304 shows only the heat-emitting portion in the picture. The signal processing apparatus 202 searches the figure for a figure based on, for example, shape and / or temperature par-35 meters associated with the figure. One typical machine vision algorithm is based on the recognition of the human head from a thermal imaging image. After the human figure is detected, the 7 signal processing equipment begins to analyze the human figure's movements and / or positions by comparing the detected positions with reference positions or reference motion vectors. If the signal processing apparatus 202 detects that human positions and / or motion vectors correlate sufficiently (similarity / correlation 5 above a predetermined threshold level) with corresponding reference reference positions and / or motion vectors, the signal processing apparatus 202 determines that the patient has crashed and informs the patient. The control unit 204 then initiates the process associated with the patient tipping event.

As mentioned above, the signal processing apparatus 202 is capable of distinguishing from the thermal camera image on the basis of the shape and / or temperature information of the human figure. The signal processing apparatus 202 analyzes the pixels of the thermal camera image, each having a temperature value. Because the human body temperature is significantly warmer (approx. 37 ° C) than ambient room air and passive objects 15 (furniture, plants, etc.), the signal processing apparatus 202 may be adapted to filter out the human body background. Signal processing equipment 202 can filter out from the image, for example, temperatures below 35-40 ° C when analyzing human body movements and postures. Signal processing equipment 202 may also filter out temperatures higher than 35-40 ° C to avoid interfering with the analysis, for example, by a stove, television or the like. After such filtering, objects other than the human body itself as well as other objects emitting heat in the temperature range can be removed from the image, as shown in Fig. 4 relative to Figs. 3A-3C. This reduces the amount of information contained in the image to be processed, which reduces the processing capacity required by the signal processing apparatus and is also well suited to the embodiment of transmitting the image to be processed over a data link to a remote point (reducing the required data throughput). The signal processing apparatus 202 may then differentiate the human body from, for example, an animal based on shape analysis.

In an environmental monitoring process performed by signal processing equipment 202, which can be performed in parallel with human motion analysis, signal processing equipment 202 analyzes an unfiltered image to monitor room conditions, or even filters out the human body image by filtering out pixels corresponding to the human body temperature. For example, signal processing equipment 202 may monitor the average room temperature, and if the temperature is found to be below the permissible lower limit (indicative of poor heating) or above the permissible upper limit (e.g., fire), the signal processing apparatus 5 202 will signal to the control unit 204. Upon receipt of a fire event signal, the control unit 204 may be configured to perform a high priority alarm, and upon receipt of a low temperature control unit 204, may be arranged to perform a low priority alarm and report information about the event.

Additionally, the signal processing apparatus 202 may be configured to recognize patient signals based on motion and position analysis. The signal processing apparatus may be configured to detect a patient's request for assistance when verifying the motion of a human subject found in the thermal imaging image to a stored reference request associated with the request for assistance. The signal processing means 202 may then notify the control unit 204 of a request for assistance, wherein the control unit 204 will perform a process associated with the request for assistance, which may include performing a predetermined alarm associated with the request for assistance. 20 The control unit may additionally notify the patient by means of an acoustic signal or some other signal that an alarm has been given. Preferably, the alarm is a high-priority alarm that requires immediate intervention by a nurse. A request for help can be predefined by waving your hand from left to right three times. Thus, the reference parameters corresponding to that path may be stored in advance in the memory unit 206. Of course, the request for assistance may be designated by another motion, but preferably a repeated motion which is not accidentally made by a person is selected as the request for assistance. This minimizes the number of false alarms. In the event of a false alarm, the signal processing equipment may also be configured to detect the movement associated with canceling the alarm. 30 If the patient discovers that he has made a false alarm based on the control unit's signaling, he may cancel the alarm with a predetermined signaling.

Figures 6A and 6B show a thermal camera image associated with an embodiment of the invention, in which the signal processing apparatus 202 is adapted 35 to determine one or more specific areas 600 associated with one or more specific modes of operation. The special area may be a portion delimited by the thermal camera image 9. The operating patterns are pre-stored in the memory unit 206, from which the signal processing apparatus 202 can read the operating models when it detects a predetermined event in a specific area that triggers the implementation of operating models. The predetermined event may be, for example, an observed shift of the human 5 character 302 into the special area 600 or a rise in temperature in the special area (e.g., stove). Operational models can define allowed events in a specific area, which eliminates the need for an alarm. If the signal processing equipment 202 detects human character movement that does not match the reference parameters associated with the allowed events, the signal processing equipment 202 will provide the control unit 10 204 with an indication of non-permitted events in the special area.

The control unit 204 then performs an operation such as, for example, an alarm via the communication means 200 and / or storing the event in the memory unit 206.

Figures 6A and 6B show an example of this embodiment. In the 15 examples, the surveillance system of the invention is located, for example, in a nursing home or hospital, and special area 600 is the patient's living area. Assume that patients do not want to leave the living room alone without the guidance of a nurse. In this case, the implementation of the allowed operating patterns may be triggered by the signal processing means detecting the transition of the human character 302 to the special area 20,600 or to the special area 600 for a predetermined time (if the patient is brought into the area). The signal processing equipment can calculate a predetermined amount of time elapsed by triggering a timer as the detected human figure moves into area 600 and adopting action patterns if the human figure is not detected to leave before the timer has expired.

When the signal processing apparatus 202 detects the implementation of the allowed operating models as the human character moves to the special area 600, it reads from the memory unit 206 the parameters corresponding to the allowed operating models. The allowed operating mode stored in the memory unit 206 may include, for example, motion parameters for moving into the special area 600 another human character 602 (Fig. 6B), which fetches 30 patients 302 out of the special area 600. clearly, so that the pixels corresponding to the human figures are connected without any background "cold" pixels between the two human figures. In order to improve the reliability of image analysis 35 in this case, a plurality of thermal cameras may be arranged to observe a specific area. This makes it possible to distinguish whether one person walks one out of the special area, or whether they walk in succession. A single thermal camera image from a particular angle may not provide this information. Thus, the signal processing apparatus 202 may analyze multiple thermal camera images from the same specific area to determine, on the basis of each thermal camera image, whether a patient is being retrieved by another human character as described above. If the signal processing apparatus 202 detects the departure of the first human figure 302 from the special area upon contact with the second human image 602 transferred to the area 602 (in each thermal camera image), this is considered a permissible mode of operation and the alarm is not performed. On the other hand, if the signal processing equipment 202 detects the departure of the first human character 302 from the special area alone, this is considered a prohibited operating pattern deviating from the allowed reference, and the signal processing equipment 202 gives the control unit 204 a signal 15 Other parameters may be included in the Permitted Policy to prevent patients from leaving the special area in pairs unless pairing is allowed.

Assuming that the patient is allowed to leave the seating area alone if he / she is in a conscious state, the system may allow the patient to leave the area 20 without alarm if the patient shows that he / she is in a conscious state by an agreed sign (e.g. In this case, if a predetermined character is detected by the signal processing system before the human character exits the special area, the exit is considered a permitted event and no alarm is performed.

25 For example, a sofa and a stove can be set up as corresponding special areas in a patient's home system. Assume that it is not good for a patient to watch television on the couch for more than a certain number of hours. When the signal processing apparatus 202 detects the movement of the human figure into the special area containing the sofa (the human figure is identified within the special area), a timer is started. When the signal processing equipment 202 detects that the patient is leaving the special area (the human figure is no longer detected in the special area), the timer is stopped and the time measured by the timer is stored in the memory unit. The control unit 204 can monitor the timer recordings on a daily basis. If the control unit 204 detects that the measured time indicated by the timer recordings exceeds a predetermined limit, the control unit 204 stores information in the memory unit 206 for watching the television that has exceeded the allowable limit on a given day.

The signal processing apparatus may also be arranged to monitor the activity of the human figure in a specific area. Activity can be determined by ana-5 lysing the human posture (sitting, lying, standing) and / or the amount of human movement. Activity can be monitored in a special area around the clock (only when a human figure is detected in the special area) or at specific times of the day. For example, the patient may be assigned a specific time to bed at which time the patient must go to bed. In this case, the signal processing apparatus 202 may be arranged to monitor the patient in a special area (a sofa in this case) beyond the time of bedtime. If the signal processing apparatus 202 detects a human figure in a specific area, the signal processing apparatus may provide the control unit 204 with a signal of the event, and the control unit 204 may transmit via the communication means 15 a message of the event (low priority alarm). In this case, the supervisor knows, for example, to call the patient and instruct the patient to go to bed. The signal processing apparatus 202 may further be arranged to monitor the position and / or activity of the human figure in a specific area. If the human figure is found to be lying down and moving infrequently, the signal processing apparatus 20 202 may determine that the patient has fallen asleep on the couch, thereby providing a control signal for this event. In this case, the control unit 204 may send a message about the event (low priority alert) via the communication means. In general, the activity of a human character can, of course, be observed in non-specific areas of the controlled area either round the clock or at specific times of the day.

The stove can be set up as a special area so that the system can monitor the stove for forgetting and warn in time before a fire starts. In this case, the introduction of allowable operating patterns may be triggered by a temperature rise detected by the signal processing means 202 above a predetermined threshold in the specific region where the cooker is located. For example, the threshold may be 100 ° C or even higher, indicating that the cooker is on. The temperature is preferably such that it can be distinguished from the thermal camera image when there is no boiler on top of the cooker or the like. This prolonged cooking does not cause an alarm. When the signal processing apparatus 202 detects a temperature above the threshold, it initiates a timer. If the temperature does not fall before the timer expires, the signal processing apparatus 12 informs the control unit 204 of a forgetting event on the hob. The control unit 204 may then be configured to alert the apartment in a specific beep when the stove is on and to trigger a timer. The signal processing apparatus 202 may continue to monitor the cooker and notify the control unit 204 if a temperature drop below the threshold is required. If the control unit 204 does not receive a notification from the signal processing device to turn off the stove within a specified time measured by the timer, the control unit 204 performs a high priority alarm via the communication means 200.

Figures 7 and 8 illustrate processes for implementing the method 10 of the invention. The processes may be implemented by computer software that controls the operation of one or more processing units. Thus, the software may be physically distributed to several entities. The signal processing means 202 and control unit 204 of Figure 2 may be implemented by one or more processors controlled by computer software. The computer-15 software may be stored on a computer-readable medium such as a memory chip, hard disk, optical storage medium, etc.

Figure 7 illustrates in general terms the process for monitoring a patient, and Figure 8 illustrates a process for performing system actions to detect a patient rollover. Referring to FIG. 7, the pro-20 process begins at step 700. The process may be initiated by the reception of a thermal video image (video) by the signal processing means 202 or by a command from the control unit 204 to initiate the process. In step 702, a received human face is searched for in a thermal cam. The search can be performed by searching the image for pixels corresponding to the human body temperature. In addition, the shape of an object formed by pixels corresponding to the human body temperature can be analyzed by comparing the shape to a stored human reference form (s). Searching can be facilitated by performing step 704, i.e., filtering out one or more temperature ranges that do not correspond to the temperature of the human body from the thermal image. Step 704 is optional, i.e. it can also be omitted from the 30 process.

When the signal processing equipment 202 has found a human figure in the thermal camera image, it compares the positions of the detected human figure with the stored reference parameters to determine the movements and positions of the character, as described above. Paragraph 708 determines whether an alarm-triggering event has been detected in human figure movements or guns. This step can be performed for event detection in signal processing equipment 202 based on analysis 13 and for decision on the need for alarm in control unit 204. If it is determined in step 708 that either no event is detected or the detected event does not require alarm, the process returns to step 706. detected, the control unit may store information about the event as it moves from step 708 to step 706. If it is determined at step 708 that an event requiring an alarm has been detected, the process goes to step 710 where the alarm is performed. Performing an alarm may include determining the type of alarm associated with the event before making the actual alarm. This allows different types of alarm 10 to be made based on the detected event. For example, a detected patient fall may not cause a high priority alarm, as described in the following example. The process returns from step 710 to step 706, where the patient's movement monitoring is continued.

Figure 8 illustrates the process of detecting a patient falling over. Step 800 moves to step 704 (or 702) after the human figure is detected in the image. At 800, the signal processing apparatus is configured to compare the positions of the human figure with one or more reference parameters associated with the crash. Step 802 determines, based on step 800, whether the movement of the detected human figure and the reference parameters associated with the fall are sufficiently correlated, i.e., whether the fall is detected. If no crash is detected, the process returns to step 800. If a human crash is detected, the process moves to step 804 where the timer is started. Step 806 begins monitoring the movement of the human figure in the signal processing apparatus 202. In step 808, it is determined whether the human figure is moving sufficiently. If it is determined that the human figure is moving sufficiently, the process returns to step 800. Otherwise, the process moves to step 710. Motion can be detected by a change in pixels depicting the temperature of the human body, and for example, a limit for this change can be detected.

Naturally, the above-described embodiments of the signal processing apparatus 202 and the control unit 204 30 may be combined to implement new embodiments. For example, the embodiment illustrating the process of FIG. 8 may be combined with the embodiment associated with the aid request of FIG. 5. Suppose a patient falls and injures his leg. In this case, he is able to move, so the process of Figure 8 may not trigger an alarm. In this case, the patient may enter a custom request for assistance request, whereby a parallel process for detecting a request for assistance request, e.g., through the functionality described in connection with Figure 5, detects a request for assistance and performs a high priority alert.

Alternatively (or additionally), the process of Figure 8 may be improved to detect patient injury, for example as a result of a fall. In this embodiment, at step 806, it is monitored whether the human figure rises from a supine position to stand or sit within the time measured by the timer. Here, the time measured by the timer can be set so high that it exceeds the time taken for the momentary unconsciousness caused by the crash. Based on the above description, the signal processing apparatus 202 is able to distinguish, by analysis and reference parameters, whether the detected human figure is in a lying, sitting or standing position. In step 808, it is determined whether the human figure is observed to move to a standing or sitting position. If so, the process returns to step 800. Otherwise, the process moves to step 710.

As another example of combining embodiments, it is described how an environmentally monitoring embodiment can be combined with an embodiment using specific domains. For example, a burglar may be detected with this embodiment. Signal processing equipment 202 may be adapted to set specific areas of the thermal camera image or images into predicted entry paths of an intruder such as an exterior door and / or windows. Signal processing equipment 202 may be configured to monitor these specific areas at certain times of the day (e.g., at night). When the signal processing apparatus 202 detects the movement of a human character in these specific areas, it provides the control unit 204 with a message about a break-in event, whereupon the control unit 204 performs an alarm. In another embodiment, for detecting a burglary, signal processing equipment 202 may be configured to monitor the number of human figures in a monitored state for a predetermined time (e.g., at night) and to compare the number of human characters detected to a predetermined number. If the number of detected human figures exceeds the allowable number-30, the signal processing apparatus 202 notifies the control unit of a corresponding event, and the control unit 204 issues a corresponding alarm. The signal processing equipment may also store the number of detected human figures in a memory unit to keep track of the number of people moving in the monitored state. In this embodiment, shape-based identification of the human figure 35 is used to prevent electronic devices or domestic animals from being interpreted as human.

15 The thermal resolution of a thermal imaging image can even be so good that it can measure human temperature. Thus, the signal processing system may be configured to measure the temperature of the human image detected in the thermal camera image by looking at the value of the pixels associated with the human image. For example, the temperature-5 background may be targeted to a position of a detected human figure known to be exposed, such as the head. In this case, the clothing will not interfere with the measurement result. If the measured temperature exceeds a threshold level, the signal processing system may provide the control unit with a signal of an event, whereby the control unit may trigger an alarm connected to the patient's high temperature. Similarly, if the measured temperature falls below a second threshold level set at the lower limit, the signal processing system may provide the control unit with a signal of an event, whereby the control unit may issue an alarm connected to the patient's too low temperature.

As mentioned at the outset, thermal cameras may not be located over the entire apartment area for cost or other reasons. For example, the washroom may be left uncovered. In this case, the signal processing apparatus 202 may be configured to start a timer when a human figure is detected exiting a designated washroom door inside the washroom. The signal processing equipment can detect this by detecting that the "warm" pixels associated with the ih-20 character disappear from the thermal camera image within the special area connected to the washroom door, and do not move outside the special area at the edge of the special area. If none of the thermal camera images detect a human being leaving the washroom (appearing in the washroom special area outside the special area edges) within the time counted by the timer 25, signal processing equipment 202 will signal to control unit 204 that event, causing control unit 204 to alert.

In connection with the above embodiment, the signal processing apparatus 202 may be configured to set the exterior door 30 of the apartment as a special area in the image to be monitored. In addition, signal processing equipment 202 may calculate the number of people in the apartment. If the signal processing equipment 202 detects that all the people charged in the apartment have left through the front door, the signal processing equipment 202 interprets this as a permitted event, so that an alarm is not performed in this case. If, on the other hand, the exit of the patient 35 from the front door is determined to be a forbidden mode of operation (e.g., at night time), signal processing equipment 202 will provide control unit 204 with a forbidden 16 exit event signal when it detects that The control unit 204 can then trigger an alarm.

It may happen that, as a result of a system reboot or similar situation, no human image is detected in any of the monitored thermal imaging images, and the system has not detected the human being removed. The system can then start the timer. If no thermal image is detected in any of the thermal imaging images within the time counted by the timer, the signal processing apparatus 202 will provide the control unit 204 with a signal of the event, and the control unit 204 may alert the supervisor.

The system may include sensors other than thermal cameras to monitor the subject. For example, a motion sensor may be installed in the washroom to provide control unit 204 with a signal upon sensing movement in the area covered by the motion sensor 15. Thus, the control unit 204 may combine the information provided by the motion sensor with the information provided by the signal processing apparatus 202. For example, if signal processing equipment 202 reports an event that a human figure has been detected exiting a washroom door in a special area and into the washroom and no patient has been detected exiting the washroom through the washroom door, control unit 204 may If the motion sensor signal indicates that there is movement in the washroom, control unit 204 determines that the patient is in the washroom and moves there, and no alarm is performed. If, on the other hand, the motion sensor has not communicated the movement within the time-limit, the control unit 204 may issue an alarm.

Although the invention has been described above with reference to the example of the accompanying drawings, it is clear that the invention is not limited thereto, but can be modified in many ways within the scope of the appended claims.

Claims (11)

A method of monitoring a person, the method comprising: one or more heat cameras placed in a space to be monitored arranged to store a heat camera image of the space to be monitored; one or more heat camera images generated by the thermal camera are analyzed (702, 706, 708) with a signal processing device by searching (702) for a human figure on the generated thermal camera image and comparing (706, 708) the parameters describing positions and / or movements of the human camera image found on the thermal camera image with corresponding stored reference parameters for reference positions and / or reference movements are performed (710) as a result of an event detected in the analysis a predetermined alarm in connection with the incident in question using data communication means functionally coupled to the signal processing device, characterized in that the human frame's auxiliary request is detected when comparing the parameters of the human camera image found and / or movement with stored parameters for reference positions and / or reference movements combined with the auxiliary request; and a predetermined alarm combined with the auxiliary request is executed as a result of the auxiliary request being detected. Method according to claim 1, characterized in that the method further comprises: detected (800, 802) in the analysis that the human figure has fallen when comparing the parameters describing the position and / or motion of the human camera image with stored parameters for reference positions and / or reference movements combined with the case; and (710) an alarm determined with the case combined is executed as a result of the discovery that the human figure has fallen. 30 3. A method according to claim 2, characterized in that the procedure further comprises: a timer (804) is started as a result of the discovery of the case; (806) the movement of the human figure is monitored after the detected case, and (808, 710) alarm is executed only when the predetermined time period measured by the timer has passed without any detected movement of the human figure. 4. A method according to any one of the preceding claims 1-3, characterized in that the analysis comprises: from the heat camera image (704), filters that deviate from a selected temperature range corresponding to the human body temperature to separate the human form from the background are filtered (704). Method according to any of the preceding claims 1-4, characterized in that the method further comprises: an event is detected when the parameters describing the positions and / or movement of the human camera image found sufficiently correlate with the stored parameters corresponding to the reference positions. and / or reference movements, and information about the occurrence of detected movements and / or positions corresponding to reference movements and / or reference positions is stored. A system for monitoring a person, the system comprising: one or more thermal cameras (100, 102) located in a space to be monitored and arranged to generate a thermal camera image of the space to be monitored; A signal processing device (20) functionally coupled to one or more thermal cameras (100, 102) and arranged to analyze a thermal camera image produced by one or more thermal cameras by looking for a human figure on the thermal camera image and comparing parameters describing the position of the thermal camera image and / or movements with reference positions and / or reference movements corresponding to stored reference parameters, and a control unit (204) which is functionally coupled to the signal processing device (20) and arranged to execute as a result of an event detected in the analysis. particular alarm in connection with the event in question using a tilly control unit (204) functionally coupled data communication unit (299), characterized in that the signal processing device (20) is configured to detect the human body's auxiliary request when it compares the parameters describing it in the heat camera vehicle n found the human body position and / or movement with stored parameters for reference positions and / or reference movements associated with the auxiliary request, when the comparison shows sufficient correlation between the human body movement and the reference movement, and the controller (204) is configured to execute one with the request request. a predetermined alarm as a result of the discovery of the request for help. 7. A system according to claim 6, characterized in that the signal processing device (202) is configured to detect in the analysis that a human has fallen as it compares the position and / or motion of the human camera image with stored parameters for the reference position. down and / or reference movements associated with the case, and that the controller (204) is configured to execute a predetermined alarm associated with the case as a result of an event associated with the case detected by the signal processing device. System according to claim 7, characterized in that the signal processing system (202) is additionally configured to start a timer as a result of the discovery of the case, to monitor the movement of the human body after the case is detected and to provide the control unit (204) with a notice of the fall event 15 only when the predetermined time period measured by the timer has passed without any movement of the human figure being detected. 9. A system according to any one of the preceding claims 6-8, characterized in that the signal processing device (20) is further configured to filter away from the thermal camera image temperatures that deviate from a waist temperature range corresponding to the human body temperature to distinguish the human figure from the background. . System according to any of the preceding claims 6-9, characterized in that the signal processing device (20) is configured to detect an event when the parameters describing the position and / or movement of the human camera image are sufficiently correlated. with the parameters corresponding to the stored reference positions and / or the reference movements, and storing the information about the event of detected movements corresponding to the reference movements. 11. Computer software product readable by computer and configured by the computer to configure the computer to perform a computer process comprising: receiving one or more thermal camera image, the thermal camera image is analyzed by looking for a human figure on the thermal camera image and comparing parameters describing it In the heat camera image, the positions of the human body found and / or movements with memory read from the reference parameters corresponding to reference positions and / or reference movements, are executed as a result of an event detected in the analysis a predetermined alarm in connection with the event in question by using the computer functionally connected data communication medium, characterized by the computer process further comprising: is detected the human-frame request for help when comparing the parameters describing the human-frame image's position and / or movement with stored parameters for reference positions and / or reference movements associated with auxiliary request; and a predetermined alarm associated with auxiliary request is executed as a result of an auxiliary request being detected.