TWI541769B - Falling down detecting systems and method - Google Patents
Falling down detecting systems and method Download PDFInfo
- Publication number
- TWI541769B TWI541769B TW104121114A TW104121114A TWI541769B TW I541769 B TWI541769 B TW I541769B TW 104121114 A TW104121114 A TW 104121114A TW 104121114 A TW104121114 A TW 104121114A TW I541769 B TWI541769 B TW I541769B
- Authority
- TW
- Taiwan
- Prior art keywords
- fall
- processing unit
- message
- detection system
- fall detection
- Prior art date
Links
Landscapes
- Alarm Systems (AREA)
- Emergency Alarm Devices (AREA)
Description
本發明係關於一種人身安全偵測系統及方法；特別是關於一種跌倒偵測系統及方法。 The present invention relates to a personal safety detection system and method; and more particularly to a fall detection system and method.
日常生活中，有許多年長者或行動不便者係單獨在家，其行動過程中，可能會因身體退化或路面不平等因素而不慎跌倒，研究顯示跌倒住院約佔意外事件的60%，且跌倒會可能造成健康惡化及產生併發症，為防止年長者或行動不便者跌倒後延誤就醫，遂發展出自動偵測跌倒的裝置。 In daily life, many elderly people or people with reduced mobility are alone at home. During the course of their actions, they may fall down due to physical deterioration or road inequality. Studies have shown that hospitalization for falls accounts for about 60% of accidents and falls. It may cause health deterioration and complications. To prevent elderly people or people with reduced mobility from delaying their medical treatment after falling, they will develop devices that automatically detect falls.
習知跌倒偵測方法須將習知跌倒偵測裝置所含的感測器(如：多軸加速器)佩戴於一受測者(如：年長者或行動不便者)身上，如：綁固於肢體等，以便於受測者跌倒時，自動發出求救訊息給相關人員，其一實施例可參酌中華民國公告第M469574號「三階段感應跌倒偵測裝置」專利案。 The conventional fall detection method is to wear a sensor (such as a multi-axis accelerator) included in a conventional fall detection device to a subject (eg, an elderly person or a person with reduced mobility), such as: Limbs, etc., to automatically send a distress message to the relevant personnel when the subject falls, an embodiment of which can refer to the Republic of China Announcement No. M469574 "Three-stage Inductive Fall Detection Device" patent case.
惟，當受測者佩戴感測器時，可能會造成受測者身心不適(如：血壓飆高等)，導致受測者不願意佩戴感測器，或者，受測者可能會忘記佩戴感測器，縱使受測者已確實佩戴感測器，亦有可能因電力中斷而使感測器失效。故，習知跌倒偵測裝置受限於感測器須被動地被受測者佩戴後才能進行偵測，其過程中會因受測者對感測器的使用習慣等因素，導致其感測效果不佳，可能造成受測者跌倒時無法即時求救。 However, when the subject wears the sensor, it may cause physical and mental discomfort (such as high blood pressure), which may cause the subject to be unwilling to wear the sensor, or the subject may forget to wear the sensor. Even if the subject has actually worn the sensor, the sensor may be disabled due to power interruption. Therefore, the conventional fall detection device is limited to the fact that the sensor must be passively worn by the subject before detecting, and the sensor may be sensed due to factors such as the user's habits of the sensor. Poor results may cause the subject to be unable to ask for help immediately when they fall.
有鑑於此，有必要改善上述先前技術的缺點，以符合實際需 求，提升其實用性。 In view of this, it is necessary to improve the shortcomings of the above prior art to meet the actual needs. Seek to improve its practicality.
本發明係提供一種跌倒偵測系統，可主動偵查受測者是否跌倒。 The invention provides a fall detection system capable of actively detecting whether a subject falls.
本發明另提供一種跌倒偵測系統，受測者無須佩戴感測器即可偵查是否跌倒。 The present invention further provides a fall detection system, in which a subject can detect whether or not he falls without wearing a sensor.
本發明再提供一種跌倒偵測方法，可主動偵查受測者是否跌倒，用以發出示警訊息。 The invention further provides a fall detection method, which can actively detect whether a subject falls or not to issue a warning message.
本發明揭示一種跌倒偵測系統，係包含：一取像元件，用以取得含有人體之連續影像；一處理單元，電性連接該取像元件，該處理單元從該連續影像中取得至少一定位點的移動狀態，比對該移動狀態與一跌倒模型，若該移動狀態與該跌倒模型相符，產生一跌倒訊息；及數個移動感測器，各移動感測器耦接該處理單元，該跌倒模型包含數個跌倒動量，各移動感測器輸出一移動訊號至該處理單元，該處理單元產生該跌倒訊息後，若該移動訊號與該跌倒動量相符，該處理單元發出該跌倒訊息。 The invention discloses a fall detection system, which comprises: an image capturing component for acquiring a continuous image containing a human body; a processing unit electrically connected to the image capturing component, the processing unit obtaining at least one positioning from the continuous image a movement state of the point, compared to the movement state and a fall model, if the movement state is consistent with the fall model, generating a fall message; and a plurality of motion sensors, each of the motion sensors coupled to the processing unit, the The fall model includes a plurality of fall momentums, and each of the mobile sensors outputs a mobile signal to the processing unit. After the processing unit generates the fall message, the processing unit issues the fall message if the mobile signal matches the fall momentum.
所述影像中的人體可具有一輪廓，該輪廓可為一軀體正常者、一軀體殘缺者或一裝設義肢者的人體輪廓。 The human body in the image may have a contour, which may be a human body contour, a body defect, or a body contour of a prosthetic person.
所述定位點可為數個，該跌倒模型可包含數種跌倒姿態，各跌倒姿態可包含數個端節點，若該數個端節點相連的輪廓符合該數個定位點相連的輪廓，該處理單元可產生該跌倒訊息。 The number of the positioning points may be several, and the falling model may include a plurality of falling postures, each of the falling postures may include a plurality of end nodes, and if the contours of the plurality of end nodes are connected to the contours of the plurality of positioning points, the processing unit This fall message can be generated.
所述數個定位點可包含一頭定位點、二手定位點及二足定位點，該數個端節點可包含一頭端節點、二手端節點及二足端節點。 The plurality of positioning points may include a positioning point, a second-hand positioning point, and a two-foot positioning point, and the number of the end nodes may include a head end node, a second-hand end node, and a two-leg end node.
所述定位點可為一個，該跌倒模型可包含數種跌倒向量，若該定位點的移動向量符合任一跌倒向量，該處理單元可產生該跌倒訊息。 The positioning point may be one, and the fall model may include several fall vectors. If the motion vector of the anchor point conforms to any fall vector, the processing unit may generate the fall message.
所述跌倒偵測系統，另包含一人機介面電性連接該處理單 元。 The fall detection system further includes a human interface electrically connecting the processing order yuan.
所述跌倒偵測系統，另包含一警報模組電性連接該處理單元。 The fall detection system further includes an alarm module electrically connected to the processing unit.
所述警報模組依據該跌倒訊息可發出一跌倒警報。 The alarm module can issue a fall alarm according to the fall message.
所述警報模組可包含一警示燈及一播報器，該警示燈及該播報器電性連接該處理單元，該警示燈可依據該跌倒警報產生一警示燈光，該播報器可依據該跌倒警報產生一播報聲音。 The alarm module can include a warning light and a broadcaster, the warning light and the broadcaster are electrically connected to the processing unit, and the warning light can generate a warning light according to the fall alarm, and the broadcaster can according to the fall alarm Produce a broadcast sound.
所述跌倒偵測系統，另包含一收發模組電性連接該處理單元。 The fall detection system further includes a transceiver module electrically connected to the processing unit.
所述收發模組可依據該跌倒訊息發出一無線訊號至一遠端平台。 The transceiver module can send a wireless signal to a remote platform according to the fall message.
所述無線訊號可經由一異質網路傳送至該遠端平台。 The wireless signal can be transmitted to the remote platform via a heterogeneous network.
所述遠端平台可為一電腦系統或一行動通訊設備。 The remote platform can be a computer system or a mobile communication device.
所述跌倒偵測系統，另包含一感測衣組，該感測衣組可設有一衣件，各移動感測器設置於該衣件。 The fall detection system further includes a sensing garment set, and the sensing garment set can be provided with a garment, and each movement sensor is disposed on the garment.
所述衣件可包含二肩部，各肩部可分別設有一移動感測器。 The garment member may include two shoulder portions, and each shoulder portion may be respectively provided with a movement sensor.
所述各跌倒動量可包含一跌倒高度及一跌倒加速度。 Each of the fall momentums may include a fall height and a fall acceleration.
本發明另揭示一種跌倒偵測方法，係包含，一判跌步驟，由一處理單元從一感測組件取得至少一定位點的移動狀態，比對該移動狀態與一跌倒模型，若該移動狀態與該跌倒模型相符，產生一跌倒訊息，否則，重新進行該判跌步驟；一呼叫步驟，由該處理單元依據該跌倒訊息發出一呼叫訊息，用以觀察受測者是否產生互動行為，若判斷為是，進行一安保步驟，若判斷為否，進行一示警步驟；該安保步驟，由該處理單元確認受測者已產生互動行為，並加以記錄；及該示警步驟，由該處理單元發出一示警訊息至一發訊模組。 The present invention further discloses a fall detection method, comprising: a step of determining a state in which a processing unit obtains a moving state of at least one positioning point from a sensing component, and compares the moving state with a falling state, if the moving state Corresponding to the fall model, a fall message is generated, otherwise, the fallback step is re-executed; in a call step, the processing unit sends a call message according to the fall message to observe whether the subject has an interactive behavior, if judging If yes, perform a security step, if the determination is no, perform an alerting step; the security step, the processing unit confirms that the subject has generated an interactive behavior, and records the same; and the warning step is issued by the processing unit The warning message is sent to a messaging module.
所述發訊模組可轉傳該示警訊息至一遠端平台。 The signaling module can forward the warning message to a remote platform.
其中該跌倒模型包含數個跌倒動量，該判跌步驟產生該跌倒訊息後，由該處理單元從數個移動感測器分別取得一移動訊號，並且比對該移動訊號與各跌倒動量，若該移動訊號與該跌倒動量相符，該處理單元執行該呼叫步驟，否則，重新進行該判跌步驟。 The fall model includes a plurality of fall momentums. After the fallback step generates the fall message, the processing unit obtains a mobile signal from the plurality of mobile sensors, and compares the mobile signal with each fall momentum. The mobile signal corresponds to the fall momentum, and the processing unit performs the calling step, otherwise, the step of resuming is repeated.
上揭跌倒偵測系統及方法之運作過程中，受測者無須佩戴感測器，即可主動偵查受測者是否跌倒，用以發出該示警訊息，可以達成「避免受測者不願意或忘記佩戴感測器等行為所導致的感測失效情事」及「在受測者跌倒時即時代其發出求救訊息」等功效。可進一步應用於居家環境及醫療院所，以減少年長者或行動不便者跌倒後延誤就醫之憾事發生。 During the operation of the fall detection system and method, the testee can take the initiative to detect whether the subject has fallen without wearing a sensor. The warning message can be used to avoid "unwilling or forgotten the subject." The effect of sensing failure caused by wearing a sensor or the like and "the message of asking for help when the subject falls." It can be further applied to the home environment and medical institutions to reduce the regrets of delays in medical treatment for the elderly or those with reduced mobility.
1‧‧‧人機介面 1‧‧‧Human Machine Interface
2‧‧‧取像元件 2‧‧‧Image taking components
3‧‧‧處理單元 3‧‧‧Processing unit
4‧‧‧警報模組 4‧‧‧Alarm module
5‧‧‧收發模組 5‧‧‧ transceiver module
6‧‧‧異質網路 6‧‧‧ Heterogeneous network
7a‧‧‧電腦系統 7a‧‧‧ computer system
7b‧‧‧行動通訊設備 7b‧‧‧Mobile communication equipment
8‧‧‧感測衣組 8‧‧‧Sense clothing group
81‧‧‧衣件 81‧‧‧clothes
811‧‧‧肩部 811‧‧‧ shoulder
82‧‧‧移動感測器 82‧‧‧Moving sensor
M‧‧‧定位點 M‧‧‧ anchor point
M1‧‧‧頭定位點 M1‧‧‧ head positioning point
M2,M3‧‧‧手定位點 M2, M3‧‧‧ hand positioning points
M4,M5‧‧‧足定位點 M4, M5‧‧‧ foot positioning point
S1‧‧‧判跌步驟 S1‧‧‧ step down
S2‧‧‧呼叫步驟 S2‧‧‧call steps
S3‧‧‧安保步驟 S3‧‧‧Security steps
S4‧‧‧示警步驟 S4‧‧‧ warning steps
第1圖：係本發明之跌倒偵測系統實施例之系統方塊圖。 Figure 1 is a block diagram of a system of an embodiment of the fall detection system of the present invention.
第2圖：係本發明之跌倒偵測方法實施例之運作流程圖。 Figure 2 is a flow chart showing the operation of the embodiment of the fall detection method of the present invention.
第3圖：係本發明之跌倒姿態示意圖。 Fig. 3 is a schematic view showing the fall posture of the present invention.
為讓本發明之上述及其他目的、特徵及優點能更明顯易懂，下文特舉本發明之較佳實施例，並配合所附圖式，作詳細說明如下：本發明全文所述之「加速度」(acceleration)，係指一物體移動時，該物體之移動速度向量對於時間之變化率，其單位可為公尺(m)/秒平方(s2)或其換算值，係本發明所屬技術領域中具有通常知識者可以理解。 The above and other objects, features and advantages of the present invention will become more <RTIgt; (Acceleration) is the rate of change of the moving velocity vector of an object with respect to time when an object moves, and its unit may be in meters (m) / second squared (s 2 ) or its converted value, which is the technology of the present invention. Those with ordinary knowledge in the field can understand.
本發明全文所述之「耦接」(coupling)，係指二電子元件之間可經由無線媒介(如：紅外線無線資訊系統傳輸、電磁或光電耦合等技術)相互傳遞訊息，係本發明所屬技術領域中具有通常知識者可以理解。 "Coupling" as used throughout the text of the present invention means that information can be transmitted between two electronic components via a wireless medium (eg, infrared wireless information system transmission, electromagnetic or optoelectronic coupling, etc.), which is a technique of the present invention. Those with ordinary knowledge in the field can understand.
請參閱第1圖所示，其係本發明之跌倒偵測系統實施例之系 統方塊圖。其中，該跌倒偵測系統實施例可包含一人機介面1、一取像元件2及一處理單元3，該人機介面1可供使用者輸入資料或輸出資料供使用者觀看，該取像元件2可取得含有人體之連續影像(如：全彩或灰階影像)，該取像元件2可電性連接該人機介面1及該處理單元3，該處理單元3可從該連續影像中取得至少一定位點的移動狀態，比對該移動狀態與一跌倒模型，若該移動狀態與該跌倒模型相符，產生一跌倒訊息。在此實施例中，該人機介面1可為含有觸控螢幕(touch panel)的電腦等，該取像元件2可為全彩或紅外線攝影機(camera)等，用以取得全彩或灰階之連續影像，該處理單元3可為微處理器(micro-processor)、特殊功能積體電路(ASIC)或數位訊號處理器(DSP)等，該處理單元3可內建一跌倒模型(如：包含數個跌倒姿態、向量、動量等)，並執行一跌倒測報程式，用以偵測一受測者是否跌倒，用以發出一警訊，惟不以此為限。 Please refer to FIG. 1 , which is a system of the fall detection system embodiment of the present invention. Unified block diagram. The embodiment of the fall detection system can include a human interface 1, an image capture component 2, and a processing unit 3, and the human interface 1 can be used by a user to input data or output data for viewing by a user. The image capturing component 2 can be electrically connected to the human interface 1 and the processing unit 3, and the processing unit 3 can obtain the continuous image from the continuous image. The movement state of at least one of the positioning points is compared to the movement state and a fall model, and if the movement state is consistent with the fall model, a fall message is generated. In this embodiment, the human interface 1 can be a computer or the like including a touch panel, and the image capturing component 2 can be a full color or an infrared camera to obtain full color or gray scale. For the continuous image, the processing unit 3 can be a micro-processor, an special function integrated circuit (ASIC) or a digital signal processor (DSP). The processing unit 3 can have a fall model built in (for example: It includes a number of fall postures, vectors, momentum, etc., and executes a fall measurement program to detect whether a subject has fallen or not to issue a warning, but not limited to this.
請再參閱第1圖所示，其中，該跌倒偵測系統實施例還可包含一警報模組4及一收發模組5，該警報模組4可電性連接該處理單元3，並可依據該跌倒訊息發出一跌倒警報，如：〝xxx已跌倒〞等；該收發模組5可電性連接該處理單元3，並可依據該跌倒訊息發出一無線訊號至一遠端平台。在此實施例中，該警報模組4可包含一警示燈及一播報器(圖未繪示)，該警示燈及該播報器可電性連接該處理單元3，該警示燈(如：閃光燈等)可依據該跌倒警報產生一警示燈光，如：該警示燈可位於大樓管理室用以發出閃爍燈光，該播報器(如：播音裝置等)可依據該跌倒警報產生一播報聲音，如：該播報器可位於大樓管理室用以發出廣播聲響；另，該收發模組5可為一無線訊號收發器，該無線訊號可經由一異質網路6(如：無線網路及有線網路之組合)傳送至一電腦系統7a或一行動通訊設備7b等遠端平台，惟不以此為限。 Please refer to FIG. 1 again, wherein the embodiment of the fall detection system further includes an alarm module 4 and a transceiver module 5, the alarm module 4 can be electrically connected to the processing unit 3, and can be The fall message sends a fall alarm, such as: 〝 xxx has fallen, etc.; the transceiver module 5 can be electrically connected to the processing unit 3, and can send a wireless signal to a remote platform according to the fall message. In this embodiment, the alarm module 4 can include a warning light and a broadcaster (not shown). The warning light and the broadcaster can be electrically connected to the processing unit 3. The warning light (such as a flash) The warning light can be generated according to the fall alarm. For example, the warning light can be located in the building management room for emitting a flashing light, and the broadcaster (such as a broadcast device, etc.) can generate a broadcast sound according to the fall alarm, such as: The broadcaster can be located in the building management room for making a broadcast sound; in addition, the transceiver module 5 can be a wireless signal transceiver, and the wireless signal can be transmitted via a heterogeneous network 6 (eg, wireless network and wired network). The combination is transmitted to a remote platform such as a computer system 7a or a mobile communication device 7b, but is not limited thereto.
請再參閱第1圖所示，其中，該跌倒偵測系統實施例亦可包 含一感測衣組8，該感測衣組8可設有一衣件81及數個移動感測器82，該衣件81可由習知衣物(如：背心、長袖衣、含帽T-shirt或工作連身服等)構成，該衣件81可穿著於一受測者身上，該衣件81可設置該移動感測器82，該移動感測器82可耦接該處理單元3，用以輔助產出該跌倒訊息。在此實施例中，該衣件81可包含二肩部811，各肩部811可分別設有一移動感測器82，惟不以此為限。 Please refer to FIG. 1 again, wherein the fall detection system embodiment may also include The sensing garment set 8 is provided with a garment 81 and a plurality of movement sensors 82. The garment 81 can be made of conventional clothing (eg, vest, long-sleeved shirt, cap-up T-shirt). Or the work piece 81 can be worn on a subject, the clothing 81 can be provided with the movement sensor 82, the movement sensor 82 can be coupled to the processing unit 3, To assist in the output of the fall message. In this embodiment, the garment 81 can include two shoulders 811, and each of the shoulders 811 can be respectively provided with a movement sensor 82, but not limited thereto.
請參閱第2圖所示，其係本發明之跌倒偵測方法實施例之運作流程圖。其中，該跌倒偵測方法實施例開始運作後，執行的步驟可包含一判跌步驟S1、一呼叫步驟S2、一安保步驟S3及一示警步驟S4，請一併參照第1圖所示，說明如下。 Please refer to FIG. 2, which is a flowchart of the operation of the embodiment of the fall detection method of the present invention. After the operation of the fall detection method is started, the steps performed may include a drop step S1, a call step S2, a security step S3, and an alert step S4, as shown in FIG. as follows.
該判跌步驟S1，可由該處理單元3從一感測組件(如第1圖所示之2及/或82)取得至少一定位點的移動狀態，比對該移動狀態與一跌倒模型，若該移動狀態與該跌倒模型相符，產生一跌倒訊息，否則，重新進行該判跌步驟S1。在此實施例中，係以該取像元件2輸出影像作為實施態樣說明，該影像可為全彩或灰階影像，惟不以此為限；該連續影像可由上述取像元件2傳送至該處理單元3，該處理單元3可從該連續影像中的其中一張影像搜尋人臉，如：使用Haar矩形特徵模型的Adaboost演算法偵測該人臉特徵，待找出人臉後，可由人臉下方依一定比例的區域定義一軀體區域，如：可由人臉區域的長、寬依據一頭身比例對應表推算其軀體比例，該人臉及軀體共同佔據的區域內可定義為一感興趣區域(interesting region)，用以取得一人身輪廓，如：利用索貝爾(Sobel)濾波器偵測輪廓邊緣，據以定義至少一定位點，如：一頭定位點、二手定位點及二足定位點等，用以追蹤後續影像中各定位點的移動狀態(如：移動方向、速度、加速度及動量等)，該移動狀態與上述跌倒模型可相互比對，若該移動狀態與該跌倒模型相符，產生該跌倒訊息。其中，其比對方式係 舉例說明如下，惟不以此為限。 In the step S1, the processing unit 3 can obtain the moving state of at least one positioning point from a sensing component (such as 2 and/or 82 shown in FIG. 1), and compare the moving state with a falling model. The movement state coincides with the fall model, and a fall message is generated. Otherwise, the fall step S1 is re-executed. In this embodiment, the image is outputted by the image capturing component 2 as an embodiment. The image may be a full color or grayscale image, but not limited thereto; the continuous image may be transmitted by the image capturing component 2 to The processing unit 3 can search for a face from one of the consecutive images, for example, an Adaboost algorithm using a Haar rectangular feature model to detect the face feature, and after the face is found, The body area defines a body area according to a certain proportion of the area. For example, the body length ratio can be calculated according to the length and width of the face area according to the ratio of the head body proportion, and the area occupied by the face and the body can be defined as an interest. The region (interesting region) is used to obtain a contour of a person, such as: using a Sobel filter to detect the edge of the contour, thereby defining at least one anchor point, such as: an anchor point, a second-hand anchor point, and a two-foot anchor point. Etc., for tracking the moving state of each positioning point in the subsequent image (such as: moving direction, speed, acceleration, momentum, etc.), the moving state and the falling model can be compared with each other, if the moving shape The fall in line with the model, resulting in the fall message. Among them, the way of comparison The examples are as follows, but not limited to this.
舉例而言，以機場通關檢驗用的紅外線攝影機所取得之連續影像為例，該人機介面1可顯示該紅外線攝影影像，用以呈現實際人體輪廓及其周圍溫度分佈情況，其中該影像顯示的人體輪廓可為一軀體正常者、一軀體殘缺者或一裝設義肢者等人體輪廓，惟紅外線攝影影像將可排除無體溫之模型(如：人體模型或人型看板等)，避免採用非人體影像所造成的誤判。因此，如第3圖之(a)所示，若該定位點為數個，則可包含一頭定位點M1、二手定位點M2、M3及二足定位點M4、M5，M1~M5中任二節點可連線構成簡易之驅體模型(如第3圖所示之(a)中所示的粗虛線)，惟不以此為限，另，該跌倒模型可包含數種跌倒姿態(如第3圖所示之(b)、(c)、(d)、(e)、(f)、(g))，各跌倒姿態包含數個端節點(如：包含一頭端節點、二手端節點及二足端節點)，若該數個端節點相連的輪廓符合該數個定位點相連的輪廓，如：兩個五角形輪廓的內角大致相同，則該處理單元可產生該跌倒訊息。 For example, taking a continuous image obtained by an infrared camera for airport customs clearance inspection, the human-machine interface 1 can display the infrared photographic image for presenting an actual human body contour and its surrounding temperature distribution, wherein the image is displayed. The outline of the human body can be a contour of a human body such as a normal body, a physically disabled person, or a prosthetic person. However, infrared photographic images can exclude models without body temperature (such as human body models or humanoid kanbans), and avoid non-human body. Misjudgment caused by images. Therefore, as shown in (a) of FIG. 3, if the number of the positioning points is several, it may include one positioning point M1, second-hand positioning points M2 and M3, and two-foot positioning points M4 and M5, and any two of M1 to M5. The connection can be made into a simple body model (such as the thick dotted line shown in (a) of Figure 3), but not limited to this. In addition, the fall model can contain several fall postures (such as the third Figure (b), (c), (d), (e), (f), (g)), each fall posture contains several end nodes (such as: including a head end node, second-hand end node and two The foot end node), if the contours of the plurality of end nodes are connected to the contours of the plurality of positioning points, for example, the inner corners of the two pentagon contours are substantially the same, the processing unit can generate the falling information.
另，該定位點亦可僅為一個(如第3圖所示，可僅含該頭定位點M)，該跌倒模型可包含數種跌倒向量(可包含移動方向及速度)，若該定位點的移動向量(可包含移動方向及速度)符合任一跌倒向量，則該處理單元可產生該跌倒訊息。 In addition, the positioning point may also be only one (as shown in FIG. 3, only the head positioning point M may be included), and the fall model may include several fall vectors (including a moving direction and a speed), if the positioning point The motion vector (which may include the direction and speed of movement) conforms to any fall vector, and the processing unit may generate the fall message.
又，該跌倒模型可包含數個跌倒動量(可包含一跌倒高度及一跌倒加速度)，各移動感測器82可感測人體之移動資訊(可包含一移動加速度及三維空間座標)並轉換為一移動訊號，並可輸出該移動訊號至該處理單元3，該處理單元3產生該跌倒訊息後，若該移動訊號與該跌倒動量相符，如：任一移動訊號中所含的移動加速度在毫秒(ms)等級的數值非為零(如：20，惟不以此為限)，且最後停留的三維空間座標接近地面，則該處理單元可發出該跌倒訊息。之後，該處理單元3可將該跌倒訊息直 接發出，或待執行該呼叫步驟S2後再決定是否發出該跌倒訊息，說明如下。 Moreover, the fall model can include a plurality of fall momentums (which can include a fall height and a fall acceleration), and each of the motion sensors 82 can sense the movement information of the human body (which can include a moving acceleration and a three-dimensional coordinate) and convert a mobile signal, and the mobile signal can be output to the processing unit 3. After the processing unit 3 generates the fall message, if the mobile signal matches the fall momentum, for example, the motion acceleration contained in any mobile signal is in milliseconds. If the value of the (ms) level is not zero (eg, 20, but not limited to this), and the last three-dimensional space coordinates are close to the ground, the processing unit can issue the fall message. Afterwards, the processing unit 3 can straighten the fall message After the call is sent, or the call is to be performed, step S2 is performed to determine whether to issue the fall message, as explained below.
該呼叫步驟S2，可由該處理單元3依據該跌倒訊息發出一呼叫訊息，用以觀察受測者是否產生互動行為，若判斷為「是」，進行該安保步驟S3，若判斷為「否」，進行該示警步驟S4。在此實施例中，該處理單元3產生該跌倒訊息後，可初步判定受測者已跌倒，此時，可進一步依據該跌倒訊息發出該呼叫訊息，如：利用播音裝置對一場所內的受測者播放〝xxx，請揮手表示安全無恙〞等呼叫訊息，使受測者依據該呼叫訊息做出反應動作(如：揮手)，該反應動作可由上述取像元件2或移動感測器82的輸出結果測得，供該處理單元3得知受測者的反應動作，用以觀察受測者是否產生互動行為，作為後續進行該安保步驟S3或該示警步驟S4之依據。 In the calling step S2, the processing unit 3 may send a call message according to the fall message to observe whether the subject has an interactive behavior. If the determination is yes, the security step S3 is performed, and if the determination is no, This warning step S4 is performed. In this embodiment, after the processing unit 3 generates the fall message, it may initially determine that the subject has fallen. At this time, the call message may be further sent according to the fall message, such as: using the broadcast device to the receiver in a place. The tester plays 〝 xxx, please wave to indicate a call message such as security, so that the test subject can respond according to the call message (eg, wave), and the reaction action can be performed by the image capture component 2 or the motion sensor 82. The output result is measured, and the processing unit 3 is informed of the reaction action of the subject to observe whether the subject generates an interactive behavior, as a basis for subsequently performing the security step S3 or the warning step S4.
該安保步驟S3，可由該處理單元3確認受測者已產生互動行為，並加以記錄。在此實施例中，該處理單元3可由上述取像元件2或移動感測器82確認受測者已產生互動行為，如：連續影像中的手定位點M2往復移動等，並儲存該互動行為發生的時間備查，惟不以此為限。 In the security step S3, the processing unit 3 can confirm that the subject has generated an interactive behavior and record it. In this embodiment, the processing unit 3 can confirm that the subject has generated an interactive behavior by the image capturing component 2 or the motion sensor 82, such as reciprocating the hand positioning point M2 in the continuous image, and storing the interaction behavior. The time of occurrence is for reference, but not limited to this.
該示警步驟S4，可由該處理單元3發出一示警訊息至一發訊模組，如：該警報模組4、收發模組5或其組合。在此實施例中，該處理單元3可產生該示警訊息，並發出至該警報模組4、收發模組5或其組合，用以產生警示燈光、播報聲音及轉傳該示警訊息至電腦系統7a、行動通訊設備7b等遠端平台，代替受測者發出求救訊息，以免貽誤救治的黃金時間。之後，可結束本發明之跌倒偵測方法實施例，亦可重新開始進行該跌倒偵測方法實施例。 In the warning step S4, the processing unit 3 can send an alarm message to a signaling module, such as the alarm module 4, the transceiver module 5, or a combination thereof. In this embodiment, the processing unit 3 can generate the warning message and send it to the alarm module 4, the transceiver module 5 or a combination thereof for generating warning lights, broadcasting sounds, and transmitting the warning messages to the computer system. 7a, mobile communication equipment 7b and other remote platforms, in lieu of the test subject to send a distress message, so as not to delay the golden time of treatment. After that, the embodiment of the fall detection method of the present invention can be ended, and the embodiment of the fall detection method can be restarted.
藉由前揭之技術手段，本發明之跌倒偵測系統及方法實施例的主要特點列舉如下：該處理單元3可從該感測組件(如：取像元件2及/或移動感測器82)取得至少一定位點的移動狀態，比對該移動狀態與一跌 倒模型，若該移動狀態與該跌倒模型相符，可產生該跌倒訊息；該處理單元3可進一步依據該跌倒訊息發出該呼叫訊息，用以觀察受測者是否產生互動行為，若判斷為「是」，可由該處理單元3確認受測者已產生互動行為，並加以記錄，若判斷為「否」，可由該處理單元3發出一示警訊息至一發訊模組(如：該警報模組4、收發模組5或其組合)。 The main features of the fall detection system and method embodiment of the present invention are as follows: The processing unit 3 can be from the sensing component (eg, the imaging component 2 and/or the motion sensor 82). Obtaining a moving state of at least one positioning point, compared to the moving state In the inverted model, if the movement state is consistent with the fall model, the fall message may be generated; the processing unit 3 may further issue the call message according to the fall message to observe whether the subject generates an interactive behavior, and if the judgment is “Yes” The processing unit 3 can confirm that the subject has generated an interactive behavior and record it. If the determination is No, the processing unit 3 can send a warning message to a signaling module (eg, the alarm module 4). , transceiver module 5 or a combination thereof).
藉此，本發明之跌倒偵測系統及方法實施例運作過程中，受測者無須佩戴感測器，即可主動偵查受測者是否跌倒，用以發出該示警訊息，可以達成「避免受測者不願意或忘記佩戴感測器等行為所導致的感測失效情事」及「在受測者跌倒時即時代其發出求救訊息」等功效。可進一步應用於居家環境及醫療院所，以減少年長者或行動不便者跌倒後延誤就醫之憾事發生。 In this way, during the operation of the embodiment of the fall detection system and method of the present invention, the subject can actively detect whether the subject has fallen without using the sensor, and to issue the warning message, thereby achieving "avoiding the test." Those who are unwilling or forget to experience the sensory failure caused by the behavior of the sensor and the "when the subject falls, that is, when the message is sent for help". It can be further applied to the home environment and medical institutions to reduce the regrets of delays in medical treatment for the elderly or those with reduced mobility.
雖然本發明已利用上述較佳實施例揭示，然其並非用以限定本發明，任何熟習此技藝者在不脫離本發明之精神和範圍之內，相對上述實施例進行各種更動與修改仍屬本發明所保護之技術範疇，因此本發明之保護範圍當視後附之申請專利範圍所界定者為準。 While the invention has been described in connection with the preferred embodiments described above, it is not intended to limit the scope of the invention. The technical scope of the invention is protected, and therefore the scope of the invention is defined by the scope of the appended claims.
1‧‧‧人機介面 1‧‧‧Human Machine Interface
2‧‧‧取像元件 2‧‧‧Image taking components
3‧‧‧處理單元 3‧‧‧Processing unit
4‧‧‧警報模組 4‧‧‧Alarm module
5‧‧‧收發模組 5‧‧‧ transceiver module
6‧‧‧異質網路 6‧‧‧ Heterogeneous network
7a‧‧‧電腦系統 7a‧‧‧ computer system
7b‧‧‧行動通訊設備 7b‧‧‧Mobile communication equipment
8‧‧‧感測衣組 8‧‧‧Sense clothing group
81‧‧‧衣件 81‧‧‧clothes
811‧‧‧肩部 811‧‧‧ shoulder
82‧‧‧移動感測器 82‧‧‧Moving sensor
Claims (19)
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| TW104121114A TWI541769B (en) | 2015-06-30 | 2015-06-30 | Falling down detecting systems and method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| TW104121114A TWI541769B (en) | 2015-06-30 | 2015-06-30 | Falling down detecting systems and method |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| TWI541769B true TWI541769B (en) | 2016-07-11 |
| TW201701238A TW201701238A (en) | 2017-01-01 |
Family
ID=56997104
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| TW104121114A TWI541769B (en) | 2015-06-30 | 2015-06-30 | Falling down detecting systems and method |
Country Status (1)
| Country | Link |
|---|---|
| TW (1) | TWI541769B (en) |
Cited By (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106781277A (en) * | 2017-01-17 | 2017-05-31 | 山西励安物联电子科技有限公司 | From calling for help warning lamp |
| TWI585716B (en) * | 2016-08-10 | 2017-06-01 | 亞洲大學 | Falling Detection Device and Method Thereof |
| TWI629455B (en) * | 2017-05-08 | 2018-07-11 | 崑山科技大學 | Hazard detection system and hazard monitoring module using relative altitude |
| US10783765B2 (en) | 2018-09-13 | 2020-09-22 | Wistron Corporation | Falling detection method and electronic system using the same |
| CN112447023A (en) * | 2019-08-29 | 2021-03-05 | 奇酷互联网络科技（深圳）有限公司 | Abnormal condition reminding method and device and storage device |
-
2015
- 2015-06-30 TW TW104121114A patent/TWI541769B/en not_active IP Right Cessation
Cited By (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| TWI585716B (en) * | 2016-08-10 | 2017-06-01 | 亞洲大學 | Falling Detection Device and Method Thereof |
| CN106781277A (en) * | 2017-01-17 | 2017-05-31 | 山西励安物联电子科技有限公司 | From calling for help warning lamp |
| CN106781277B (en) * | 2017-01-17 | 2023-10-27 | 刘海超 | Self-help calling warning lamp |
| TWI629455B (en) * | 2017-05-08 | 2018-07-11 | 崑山科技大學 | Hazard detection system and hazard monitoring module using relative altitude |
| US10783765B2 (en) | 2018-09-13 | 2020-09-22 | Wistron Corporation | Falling detection method and electronic system using the same |
| CN112447023A (en) * | 2019-08-29 | 2021-03-05 | 奇酷互联网络科技（深圳）有限公司 | Abnormal condition reminding method and device and storage device |
Also Published As
| Publication number | Publication date |
|---|---|
| TW201701238A (en) | 2017-01-01 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| TWI541769B (en) | Falling down detecting systems and method | |
| JP7138931B2 (en) | Posture analysis device, posture analysis method, and program | |
| US10825314B2 (en) | Baby monitor | |
| Stone et al. | Fall detection in homes of older adults using the Microsoft Kinect | |
| Ozcan et al. | Automatic fall detection and activity classification by a wearable embedded smart camera | |
| WO2015133195A1 (en) | Information processing device, information processing method, and program | |
| CN107257651A (en) | The scene detection of medical monitoring | |
| US10959646B2 (en) | Image detection method and image detection device for determining position of user | |
| JP6086468B2 (en) | Subject monitoring system | |
| TWI384383B (en) | Apparatus and method for recognizing gaze | |
| JP2008541650A (en) | Monitoring method and apparatus | |
| Ciabattoni et al. | Fall detection system by using ambient intelligence and mobile robots | |
| Bansal et al. | Elderly people fall detection system using skeleton tracking and recognition | |
| CN108846996B (en) | Tumble detection system and method | |
| GB2593931A (en) | Person monitoring system and method | |
| TW200933535A (en) | Intelligent nursing system | |
| JP6822326B2 (en) | Watching support system and its control method | |
| JP6870514B2 (en) | Watching support system and its control method | |
| Safarzadeh et al. | Real-time fall detection and alert system using pose estimation | |
| JP5870230B1 (en) | Watch device, watch method and watch program | |
| US11087157B2 (en) | Image detection method and image detection device utilizing dual analysis | |
| US11257246B2 (en) | Image detection method and image detection device for selecting representative image of user | |
| KR102544147B1 (en) | Image Analysis based One Person Fall Detection System and Method | |
| US11114198B2 (en) | Monitoring an individual's condition based on models generated from e-textile based clothing | |
| TWM605852U (en) | System for detecting body temperature, face and mask |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| MM4A | Annulment or lapse of patent due to non-payment of fees |