TW201635247A - Abnormality determination system based on sound and image changes - Google Patents

Abnormality determination system based on sound and image changes Download PDF

Info

Publication number
TW201635247A
TW201635247A TW104109216A TW104109216A TW201635247A TW 201635247 A TW201635247 A TW 201635247A TW 104109216 A TW104109216 A TW 104109216A TW 104109216 A TW104109216 A TW 104109216A TW 201635247 A TW201635247 A TW 201635247A
Authority
TW
Taiwan
Prior art keywords
sound
image
abnormality
area
decibel
Prior art date
Application number
TW104109216A
Other languages
Chinese (zh)
Inventor
hong-ming Liu
Original Assignee
Xinyi Dev Co Ltd
Xinyi Dev Shanghai Co Ltd
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Xinyi Dev Co Ltd, Xinyi Dev Shanghai Co Ltd filed Critical Xinyi Dev Co Ltd
Priority to TW104109216A priority Critical patent/TW201635247A/en
Publication of TW201635247A publication Critical patent/TW201635247A/en

Links

Landscapes

  • Alarm Systems (AREA)

Abstract

The present invention relates to an abnormality determination system based on sound and image changes for determining whether an abnormality occurs in a region. The abnormality determination system mainly includes a camera device, a sound receiving device, and an abnormality determination device. The camera device shoots the region and generates a regional image. The sound receiving device receives sound from the region and generates a regional sound signal of the region. As such, the abnormality determination device can determine whether an abnormality occurs in respect of the received regional image in order to determine whether the image is abnormal or normal, and determine whether an abnormality occurs in respect of the received regional sound signal of the region in order to determine whether the sound is abnormal or normal. When determining the image is abnormal and/or the sound is abnormal, the abnormality determination device issues an abnormality determination signal.

Description

基於聲音與影像變化的異常判斷系統 Abnormal judgment system based on sound and image changes

本發明係關於一種異常判斷系統,尤其是基於聲音與影像變化的異常判斷系統。 The present invention relates to an abnormality judging system, and more particularly to an abnormality judging system based on sound and image changes.

隨著家庭收入的提高和人民的逐漸富裕,家中存放財務和貴重物品的數量和價值都在不斷提高,使得家居安全在現代社會也越來越重要。雖說貴重物品可以存入銀行保險箱,但是,那畢竟不方便取用,尤其是對於女性的首飾和家裡常用的存摺、銀行提款卡、現金之類的存放,而且存入保險箱也會讓貴重物品失去在家庭中的裝飾作用。 With the increase in household income and the gradual affluence of the people, the quantity and value of financial and valuable items stored in the home are constantly increasing, making home security more and more important in modern society. Although valuables can be stored in the bank safe, it is not convenient to use, especially for women's jewelry and household passbooks, bank ATM cards, cash and other storage, and depositing safes will also make valuables Lose the decorative effect in the family.

現在人收入提高後,外出旅遊的人也在增加,於是就會形成很多居室在一定時間內無人居住和無人看管的現狀,這種情況就給盜竊人員找到了可乘之機。因此,安全的生活和居住環境越來越引起人們的關注,防盜門、防盜窗等安防設施都應運而生。 Now that people's income has increased, the number of people who travel abroad has increased. As a result, many homes are unoccupied and unattended for a certain period of time. This situation has given the thieves an opportunity to take advantage of. Therefore, the safe living and living environment has attracted more and more people's attention, and security facilities such as security doors and security windows have emerged.

近年來一種新型的保全設施一影像監控系統,以其智慧化和友善化而備受用戶青睞,在各行業得到了廣泛的應用。不過,傳統影像監控系統只能提供監控區域的圖像,使得對監控區域的監控、對目標物件的識別、分析、辨別、跟蹤等完全依靠人來進行。 In recent years, a new type of security facilities-image surveillance system has been favored by users because of its wisdom and friendliness, and has been widely used in various industries. However, the traditional image monitoring system can only provide images of the monitoring area, so that the monitoring of the monitoring area, the identification, analysis, identification, and tracking of the target object are completely dependent on the person.

為了改善上述問題,智慧影像監控具備了自動、智慧的圖像分析能力,可在不需要人為幹預的情況下,通過對攝像機拍攝的圖像的自 動分析,可以識別、分辨場景中的動態目標,取得目標的大小、數量、方向、速度等資訊,能夠在異常情況發生的時候以最快和最佳的方式做出告警、錄影、跟蹤等反應,從而大大節約人員數量、降低勞動強度、最大限度的降低誤報和漏報現象、減少對存放裝置的投資等。然而,傳統的智慧影像監控對光線十分敏感,很容易因為光影或光線不足下發生誤報和漏報。 In order to improve the above problems, intelligent image monitoring has an automatic and intelligent image analysis capability, which can automatically capture images taken by the camera without human intervention. Dynamic analysis can identify and distinguish dynamic targets in the scene, obtain information such as size, quantity, direction, speed, etc., and can generate alarms, videos, tracking, etc. in the fastest and best way when abnormal conditions occur. Therefore, the number of personnel is greatly reduced, the labor intensity is reduced, the false positives and false negatives are minimized, and the investment in the storage device is reduced. However, traditional intelligent image monitoring is very sensitive to light, and it is easy to cause false positives and false negatives due to lack of light or light.

為了改善這個問題,已有廠商利用紅外線遮蔽或感應壓力來加強誤報和漏報。但,紅外線感應還是壓感晶片,都有一定的不足之處。紅外線感應方式只有對射的架設方式,所以有被跨越的可能,無法做到整個空間的感應。紅外線具有的水溶性而被吸收的情況,因此在潮氣過高的地方很容易受到影響,而有誤判的情況。另外,壓感晶片方式一方面也存在紅外線的弊端,那就是無法對空間進行感應,同時需要大量的貼片,無形中在達到近似效果的時候需要更多的成本投入和實施難度。 In order to improve this problem, manufacturers have used infrared shielding or induced pressure to enhance false positives and false negatives. However, infrared sensing or pressure sensitive wafers have certain deficiencies. The infrared sensing method has only the way of erecting the radiation, so there is a possibility of being crossed, and it is impossible to sense the entire space. Since infrared rays are water-soluble and absorbed, they are easily affected in places where moisture is too high, and there is a case of misjudgment. In addition, the pressure-sensitive wafer method also has the drawback of infrared rays on the one hand, that is, it is impossible to sense the space, and at the same time, a large number of patches are required, which inevitably requires more cost investment and implementation difficulty when the approximate effect is achieved.

本發明之主要目的在提供一種基於聲音與影像變化的異常判斷系統,其藉著聲音與影像同時都出現超過某種程度的變化時,才發出異常判斷訊號,避免單獨影像或聲音異常判斷過高的誤判率,也避免增加過多軟硬體成本來降低誤判率。 The main object of the present invention is to provide an abnormality judging system based on sound and image changes, which emits an abnormality judgment signal when sound and image appear more than a certain degree at the same time, so as to avoid excessive judgment of individual images or sounds. The false positive rate also avoids increasing the cost of soft and hardware to reduce the false positive rate.

基於上述目的,本發明基於聲音與影像變化的異常判斷系統,用以判斷區域是否出現異常。異常判斷系統主要包攝影裝置、聲音接收裝置、異常判斷裝置。攝影裝置對該區域進行拍攝並產出區域影像,而聲音接收裝置則對該區域進行收音並產出區域區域聲音訊號。如此,異常判斷裝置即可針對所接收到的區域影像判斷出是否有異常,而判斷出影像 判斷異常或影像判斷正常,並針對所接收到的區域區域聲音訊號判斷出是否有異常,而判斷出聲音判斷異常或聲音判斷正常。當異常判斷裝置判斷出影像判斷異常或/和聲音判斷異常時,發出異常判斷訊號。 Based on the above object, the present invention is based on an abnormality determination system for sound and image changes for determining whether an abnormality occurs in a region. The abnormality determination system mainly includes a photographing device, a sound receiving device, and an abnormality determining device. The photographing device photographs the area and produces an area image, and the sound receiving device picks up the area and produces an area area sound signal. In this way, the abnormality determining device can determine whether there is an abnormality for the received region image, and determine the image. It is judged that the abnormality or the image judgment is normal, and it is judged whether there is an abnormality with respect to the received regional area sound signal, and it is judged that the sound judgment is abnormal or the sound judgment is normal. When the abnormality determining device determines that the image determination is abnormal or/and the sound determination is abnormal, an abnormality determination signal is issued.

關於本發明之優點與精神可以藉由以下的發明詳述及所附圖式得到進一步的瞭解。 The advantages and spirit of the present invention will be further understood from the following detailed description of the invention.

10‧‧‧攝影裝置 10‧‧‧Photographing device

20‧‧‧聲音接收裝置 20‧‧‧Sound receiver

30‧‧‧異常判斷裝置 30‧‧‧Abnormal judgment device

401、402‧‧‧區域影像 401, 402‧‧‧ regional imagery

501、502‧‧‧區域聲音訊號 501, 502‧‧‧ regional audio signals

601‧‧‧分貝對照表 601‧‧ ‧ decibel table

602‧‧‧距離分貝對照表 602‧‧‧ distance decibel table

第一A~一B圖為本發明基於聲音與影像變化的異常判斷系統之示意圖。 The first A~B diagram is a schematic diagram of an abnormality judgment system based on sound and image changes of the present invention.

第二A~二B圖為本發明異常判斷裝置之運作模式示意圖。 The second A to B diagrams are schematic diagrams showing the operation mode of the abnormality determining apparatus of the present invention.

第三A~三B圖為本發明異常判斷之示意圖。 The third A~3B diagram is a schematic diagram of the abnormal judgment of the present invention.

第四A~四B圖為本發明聲音判斷異常之示意圖。 The fourth A to the fourth B diagrams are schematic diagrams of the abnormal sound judgment of the present invention.

請參閱第一A~一B圖,第一A~一B圖為本發明基於聲音與影像變化的異常判斷系統之示意圖。如第一A圖所示,本發明異常判斷系統主要包含攝影裝置10、聲音接收裝置20、異常判斷裝置30。攝影裝置10對區域進行拍攝並產出區域影像,而聲音接收裝置20則對該區域進行收音並產出區域區域聲音訊號。 Please refer to the first A~B diagram. The first A~B diagram is a schematic diagram of the abnormality judgment system based on sound and image change in the present invention. As shown in FIG. A, the abnormality determination system of the present invention mainly includes a photographing device 10, a sound receiving device 20, and an abnormality determining device 30. The photographing device 10 photographs the area and produces an area image, and the sound receiving device 20 picks up the area and produces an area area sound signal.

舉例來說,如果本發明基於聲音與影像變化的異常判斷系統所監控的區域為社區的牆壁,而監控是否有宵小翻越牆壁,入侵社區時,攝影裝置10與聲音接收裝置20可採用「影音同步」的機種,或在原有監視器旁加裝一個現場收音「麥克風」。為了監控翻牆與翻落地面等行為,攝影裝置10對準牆壁,並續監控牆壁上方的影像變化,同時聲音接收裝置20則可採用麥克風陣列或指向麥克風針對牆壁與鄰近牆壁的地面進行聲音監 控。 For example, if the area monitored by the abnormality judging system based on the sound and image changes of the present invention is the wall of the community, and the monitoring of whether there is a small over-the-wall and invading the community, the photographing apparatus 10 and the sound receiving apparatus 20 can adopt "video synchronization". For the model, or add a live radio "microphone" next to the original monitor. In order to monitor the behavior of overturning the wall and falling over the ground, the photographic device 10 is aligned with the wall and continuously monitors the image changes above the wall, while the sound receiving device 20 can use the microphone array or the pointing microphone to perform sound monitoring on the wall and the wall adjacent to the wall. control.

簡單來說,異常判斷裝置30可針對所接收到的區域影像判斷 出是否有異常,而判斷出影像判斷異常或影像判斷正常,並針對所接收到的區域區域聲音訊號判斷出是否有異常,而判斷出聲音判斷異常或聲音判斷正常。當異常判斷裝置判斷出影像判斷異常或/和聲音判斷異常時,發出異常判斷訊號給監控中心或相關人員,也可以利用事件發生所在地的擴音機發出警報或點亮警示燈,嚇退入侵者。 Briefly, the abnormality determining device 30 can determine the received region image If there is an abnormality, it is judged that the image judgment is abnormal or the image judgment is normal, and it is judged whether there is an abnormality with respect to the received area area sound signal, and it is judged that the sound judgment is abnormal or the sound judgment is normal. When the abnormality determining device determines that the image determination is abnormal or/and the sound determination is abnormal, an abnormality determination signal is sent to the monitoring center or related personnel, and the amplifier of the place where the event occurs may be used to issue an alarm or light a warning light to scare off the intruder.

如第一B圖所示,為了讓使用者更有彈性的使用本發明系 統,除了監控影像與聲音都出現異常外,也分別針對只有影像判斷異常、聲音判斷異常、以及影像判斷異常或聲音判斷異常個別發生異常時,也發出異常判斷訊號給監控中心或相關人員,同樣也可以利用事件發生所在地的擴音機發出警報或點亮警示燈,嚇退入侵者。 As shown in Figure B, in order to allow users to use the system more flexibly In addition to the abnormality of monitoring images and sounds, it also sends an abnormality judgment signal to the monitoring center or related personnel for the case where only the image judgment is abnormal, the sound judgment is abnormal, and the image judgment is abnormal or the sound judgment is abnormal. You can also use the amplifier at the location where the event occurred to sound an alarm or illuminate a warning light to scare off the intruder.

不過,有關於上述異常判斷機制還需要進一步說明。 However, there is a need for further explanation regarding the above abnormality judgment mechanism.

請參閱第二A~二B圖,第二A~二B圖為本發明異常判斷裝置之運作模式示意圖。在本發明異常判斷系統中,為了配合上述多種異常判斷機置,本發明異常判斷裝置30需要先判斷出影像判斷異常或聲音判斷異常。如第二A圖所示,本發明異常判斷裝置30在所接收到的該區域影像的第一變動程度超過第一變動比較值而為該影像判斷異常、以及區域區域聲音訊號的第二變動程度超過第一變動比較值而為聲音判斷異常時,發出異常判斷訊號。不過,異常判斷裝置30也可以不需要先判斷出影像判斷異常或聲音判斷異常,而可以在異常判斷裝置30所接收到的區域影像的第一變動程度超過第一變動比較值、以及區域區域聲音訊號的第二變動程度超過第 二變動比較值時,就發出異常判斷訊號,如第二B圖所示。底下,將分別針對影像判斷異常、聲音判斷異常更具體說明。 Please refer to the second A~B diagram, and the second A~B diagram is a schematic diagram of the operation mode of the abnormality determining apparatus of the present invention. In the abnormality determination system of the present invention, in order to cooperate with the above-described various abnormality determination devices, the abnormality determination device 30 of the present invention needs to first determine an abnormality in image determination or abnormality in sound determination. As shown in FIG. 2A, the abnormality determining device 30 of the present invention determines that the image is abnormal and the second degree of variation of the regional region sound signal is that the first degree of variation of the received region image exceeds the first variation comparison value. When the first change comparison value is exceeded and the sound is judged to be abnormal, an abnormality determination signal is issued. However, the abnormality determining device 30 does not need to first determine the image determination abnormality or the sound determination abnormality, and the first variation degree of the region image received by the abnormality determining device 30 may exceed the first variation comparison value and the regional region sound. The second change in the signal is greater than the first When the comparison value is changed, an abnormality judgment signal is issued, as shown in the second B diagram. Underneath, the image judgment abnormality and the sound judgment abnormality will be described in more detail.

請參閱第三A~三B圖,第三A~三B圖為本發明異常判斷之示 意圖。如第三A圖所示,異常判斷裝置30先在第一時間進行拍攝,然後再第二時間也進行拍攝,接著根據在第一時間所接收到的區域影像401、以及在第二時間所接收到的區域影像402來比較出差異影像。不過,區域影像401也可以是在預訂時間內所接收到的區域影像平均值。換言之,區域影像401可以是事先準備好的影像,也可以設定每個整點所拍攝到影像,以便可以自動地隨著白天或晚上等等外部環境光線變化而跟著改變基準影像,不需要人為指定哪個時間點為基準影像。這其中,如果攝影裝置10屬於旋轉式的,所拍得的區域影像可能缺乏相對性,因此可以利用影像比對與對齊技術,找到第一時間所接收到的區域影像401、以及在該第二時間所接收到的區域影像402為對齊,以便補償攝影裝置10在移動攝影模式下所造成的變動。 Please refer to the third A~3B diagram, and the third A~3B diagram is the indication of the abnormal judgment of the present invention. intention. As shown in FIG. 3A, the abnormality determining device 30 first performs shooting at the first time, and then performs shooting at the second time, and then receives the area image 401 received at the first time and received at the second time. The resulting region image 402 is compared to the difference image. However, the regional image 401 may also be the average of the regional images received during the reservation time. In other words, the area image 401 can be an image prepared in advance, or an image can be set for each hour, so that the reference image can be automatically changed with the change of the external environment such as day or night, and no artificial designation is required. Which time point is the reference image. Among them, if the photographing device 10 is of a rotary type, the captured area image may lack relativity, so the image matching and alignment technique can be used to find the area image 401 received at the first time, and in the second The area image 402 received at the time is aligned to compensate for variations caused by the photographing apparatus 10 in the moving photographing mode.

舉例來說,區域影像401原本只拍到攝牆壁與附近的草木, 但區域影像402則在原本的牆壁上出現大約50X200公分的異常物體。在這種情況下,異常判斷裝置30在第一時間與第二時間內所比較出的屬於第一變動程度的差異影像,只要超過第一變動比較值時,則判斷出影像判斷異常。 For example, the area image 401 originally only photographed the wall and the nearby vegetation. However, the area image 402 has an abnormal object of about 50 x 200 cm on the original wall. In this case, the difference image belonging to the first degree of change compared between the first time and the second time by the abnormality determining device 30 determines that the image determination is abnormal when the first variation comparison value is exceeded.

舉例來說,如果第一變動比較值為比例值時,則可依據異常 物體佔據整個區域影像的比例值作為影像判斷異常的標準(例如,異常物體佔據整個區域影像超過10%)。為了達到這個效果,異常判斷裝置30首先利用景深地圖(depth map)技術判斷出差異影像中出現異常物的異常物概率分 佈,而找出足夠大的差異,避免受到如落葉等雜訊干擾,導致誤判。然後,依據異物概率分佈,讓異常判斷裝置30利用背景相減技術從第二時間所接收到的區域影像402分析出異常物體。 For example, if the first change comparison value is a proportional value, then the exception can be based on The proportion of the object occupying the entire area image is used as a criterion for image anomaly (for example, an anomalous object occupies more than 10% of the entire area image). In order to achieve this effect, the abnormality determining device 30 first uses the depth map technique to determine the probability of the abnormality of the abnormal object in the difference image. Cloth, and find a large enough difference to avoid interference with noise such as leaves, leading to misjudgment. Then, based on the foreign matter probability distribution, the abnormality determining means 30 analyzes the abnormal object from the area image 402 received at the second time by the background subtraction technique.

當區域影像中異常物體所佔據的範圍超過屬於第一變動比 較值的比例值(例如10%)時,以及區域區域聲音訊號的第二變動程度超過該第二變動比較值時,發出異常判斷訊號。當然,如果只有區域影像中異常物體所佔據的範圍超過10%也可以發出判斷訊號。 When the anomalous object in the area image occupies more than the first variation ratio When the ratio value (for example, 10%) is compared, and the second degree of change of the regional area sound signal exceeds the second variation comparison value, an abnormality determination signal is issued. Of course, if only the range of the anomalous objects in the area image is more than 10%, a judgment signal can be issued.

聲音判斷異常的部份原理上大致上差不多。三B圖所示,聲 音接收裝置20先在第一時間進行收音,然後再第二時間也進行收音,接著根據在第一時間所接收到的區域聲音訊號501、以及在第二時間所接收到的區域聲音訊號502來比較出差異分貝。 The part of the sound judgment abnormality is roughly the same in principle. As shown in Figure 3B, the sound The tone receiving device 20 first receives the sound at the first time, and then performs the sound collection at the second time, and then according to the regional sound signal 501 received at the first time and the regional sound signal 502 received at the second time. Compare the difference decibels.

接著,異常判斷裝置30根據在第一時間所接收到的區域分貝 值501、以及在第二時間所接收到的區域分貝值502來比較出差異分貝,其中第一時間早於該第二時間。同樣地,第一時間所接收到的區域分貝值501可用在預訂時間內所接收到的區域分貝值來替代。 Next, the abnormality determining means 30 is based on the area decibel received at the first time. The value 501, and the region decibel value 502 received at the second time, compares the difference decibels, wherein the first time is earlier than the second time. Similarly, the region decibel value 501 received at the first time can be replaced with the region decibel value received during the reservation time.

如此,當區域影像的該第一變動程度超過第一變動比較值 時,以及屬於第二變動程度的差異分貝的超過第二變動比較值時,異常判斷裝置30發出異常判斷訊號。在進行聲音異常判斷時,可選擇實體物(例如宵小)可發出的聲音頻段內的該區域分貝值進行比較出該差異分貝,排除被非監控物件(例如極低頻或超高音訊)所發出的聲音干擾,而該聲音頻段為20Hz~4KHz。 Thus, when the first variation of the regional image exceeds the first variation comparison value When the difference decibel belonging to the second degree of variation exceeds the second variation comparison value, the abnormality determining means 30 issues an abnormality determination signal. In the case of abnormal sound judgment, the decibel value of the region in the sound frequency band that can be emitted by the physical object (for example, small) can be selected to compare the differential decibel to exclude the unmarked object (for example, extremely low frequency or super high audio). The sound interferes, and the sound frequency band is 20 Hz to 4 kHz.

需特別注意的是,聲音接收裝置20可採用分貝計,直接針對 所測到的聲音產出數位訊號的分貝值。如果是利用一般麥克風,要產出數位訊號的分貝值則需要有轉換機制。 It should be noted that the sound receiving device 20 can be directly measured against the decibel meter. The measured sound produces a decibel value of the digital signal. If a general microphone is used, a conversion mechanism is required to produce a decibel value of the digital signal.

請參閱第四A~四B圖,第四A~四B圖為本發明聲音判斷異常 之示意圖。如第四A圖所示,異常判斷裝置30將所接收到的區域區域聲音訊號處理成數位訊號後,並基於分貝對照表601而產出區域分貝值。這其中,分貝對照表601包括有複數個數位訊號及匹配數位訊號的分貝值。簡單講,分貝對照表可將單純的區域聲音訊號轉換成分貝值。 Please refer to the fourth A~4B picture, and the fourth A~4B picture is the abnormal sound judgment of the present invention. Schematic diagram. As shown in FIG. 4A, the abnormality determining device 30 processes the received regional region audio signal into a digital signal, and generates a regional decibel value based on the decibel comparison table 601. Among them, the decibel comparison table 601 includes a decibel value of a plurality of digital signals and matching digital signals. Simply put, the decibel table can convert a simple regional sound signal into a shell value.

不過,在實際的測試中,需要測定並非異常判斷裝置30所在 地的分貝值,而是預估鄰近聲音接收裝置20的一特定地點的分貝值,因此必須要根據實際或預估的距離值,補償距離遠近造成的能量衰減損耗,來確定預估分貝值。 However, in the actual test, it is necessary to determine that the abnormality determining device 30 is not located. The decibel value of the ground is estimated as the decibel value of a specific location of the adjacent sound receiving device 20. Therefore, it is necessary to compensate the energy attenuation loss caused by the distance according to the actual or estimated distance value to determine the estimated decibel value.

如第四B圖所示,在轉換出數位訊號的分貝值的過程中,在 區域區域聲音訊號處理成數位訊號後,還包含有音源能量衰減計算程式,依據距離,基於距離分貝對照表602而產出區域分貝值。其中,距離分貝對照表602包括有複數個數位訊號及匹配數位訊號與距離的分貝值。分貝對照表602是可通過音源能量衰減計算方程式,或通過事先反復實驗後,將所得的預估分貝值存儲在該距離值與預估分貝值對照表中,以便供實際使用時,根據使用者輸入的距離,來換算出特定位置的預估分貝值。由於該音源能量衰減計算方程式,是一般自然原理的運用,並非本發明所專有,且已廣泛應用在各種領域內,早已為大眾所周知,故在此不予以贅述。 As shown in the fourth B diagram, in the process of converting the decibel value of the digital signal, After the regional area sound signal is processed into a digital signal, the sound source energy attenuation calculation program is further included, and the area decibel value is generated based on the distance based on the distance decibel comparison table 602. The distance decibel table 602 includes a plurality of digital signals and a decibel value matching the digital signal and the distance. The decibel comparison table 602 can calculate the equation by the sound source energy attenuation, or by repeatedly experimenting in advance, and storing the obtained estimated decibel value in the distance value and the estimated decibel value comparison table, so as to be used for actual use, according to the user. Enter the distance to convert the estimated decibel value for a particular location. Since the equation for calculating the energy attenuation of the sound source is an application of the general natural principle, it is not exclusive to the present invention, and has been widely used in various fields, and has long been well known, and therefore will not be described herein.

在以上描述中,所需的距離數值可利用影像技術推算得知, 不一定需要使用者手動輸入聲音接收裝置20與所監控場所(例如牆壁)之間 的距離。也就是說,攝影裝置10與聲音接收裝置20為一體或彼此實體有接觸的情況下,上述距離可為攝影距離。為了得出攝影距離,可利用景深地圖(depth map)技術所判斷出異常物體與攝影裝置10之間的距離。 In the above description, the required distance value can be estimated by using image technology. It is not necessary for the user to manually input between the sound receiving device 20 and the monitored place (such as a wall) the distance. That is to say, in the case where the photographing device 10 and the sound receiving device 20 are integrated or physically in contact with each other, the above distance may be a photographing distance. In order to derive the photographing distance, the distance between the abnormal object and the photographing device 10 can be judged by the depth map technique.

為了區別分貝數是經過距離校正過,如果在聲音接收裝置20 所處位置的噪音所測得的分貝值為60dB,而可在該顯示器上可顯示為60dB/0m;若是測得特定位置(例如牆壁)距離該聲音接收裝置20為20M與預估分貝值為70dB時,其顯示該特定位置的預估分貝值,則可根據對所測得分貝值及輸入之距離值進行處理,以顯示出預估分貝值70dB/20m;或者是同時將各個特定位置(聲音接收裝置20至噪音源範圍內)的預估分貝值都在該顯示器上顯示出來,如60dB/0m、61dB/2m,65dB/10m…70dB/20m(但是,該預估分貝值只限於聲音接收裝置20與聲源的直線路徑及其鄰近區域,若偏差太遠,則應視聲源及聲音接收裝置20的具體位置而定)。 In order to distinguish the decibel number, it is corrected by the distance if it is at the sound receiving device 20 The measured position has a decibel value of 60 dB, which can be displayed as 60 dB/0 m on the display; if the specific position (eg, wall) is measured, the sound receiving device 20 is 20 M and the estimated decibel value is 20 dB. At 70dB, it displays the estimated decibel value of the specific position, and can process the measured value of the measured value and the input distance value to display the estimated decibel value of 70dB/20m; or at the same time each specific position ( The estimated decibel values of the sound receiving device 20 to the noise source range are displayed on the display, such as 60dB/0m, 61dB/2m, 65dB/10m...70dB/20m (however, the estimated decibel value is limited to sound). The linear path of the receiving device 20 and the sound source and its adjacent area, if the deviation is too far, depends on the specific location of the sound source and the sound receiving device 20.

藉由以上較佳具體實施例之詳述,係希望能更加清楚描述本 發明之特徵與精神,而並非以上述所揭露的較佳具體實施例來對本發明之範疇加以限制。相反地,其目的是希望能涵蓋各種改變及具相等性的安排於本發明所欲申請之專利範圍的範疇內。 With the above detailed description of the preferred embodiments, it is desirable to describe this more clearly. The invention is not limited by the specific embodiments disclosed herein. On the contrary, the intention is to cover various modifications and equivalents within the scope of the invention as claimed.

10‧‧‧攝影裝置 10‧‧‧Photographing device

20‧‧‧聲音接收裝置 20‧‧‧Sound receiver

30‧‧‧異常判斷裝置 30‧‧‧Abnormal judgment device

Claims (19)

一種基於聲音與影像變化的異常判斷系統,用以判斷一區域是否出現異常,該異常判斷系統包含:一攝影裝置,用以對該區域進行拍攝,並產出一區域影像;一聲音接收裝置,用以對該區域進行收音,並產出一區域區域聲音訊號;以及一異常判斷裝置,針對所接收到的該區域影像判斷出是否有異常,而判斷出一影像判斷異常或一影像判斷正常,並針對所接收到的該區域區域聲音訊號判斷出是否有異常,而判斷出一聲音判斷異常或一聲音判斷正常;其中,當該異常判斷裝置判斷出該影像判斷異常或/和該聲音判斷異常時,發出一異常判斷訊號。 An abnormality judging system based on sound and image changes for judging whether an area has an abnormality, the abnormality judging system comprising: a photographing device for photographing the area and generating an area image; a sound receiving device, The method is used for collecting the sound of the area, and generating an area area sound signal; and an abnormality determining device, determining whether an abnormality is detected for the received image of the area, and determining that an image is abnormal or an image is judged to be normal. And determining whether there is an abnormality in the received regional area audio signal, and determining that a sound determination is abnormal or a sound determination is normal; wherein the abnormality determining device determines that the image determination is abnormal or/and the sound is abnormally determined At the time, an abnormal judgment signal is issued. 如申請專利範圍第1項所述之基於聲音與影像變化的異常判斷系統,其中該異常判斷裝置在所接收到的該區域影像的一第一變動程度超過一第一變動比較值而為該影像判斷異常、以及該區域區域聲音訊號的一第二變動程度超過一第一變動比較值而為該聲音判斷異常時,發出該異常判斷訊號。 The abnormality judging system based on sound and image change according to claim 1, wherein the abnormality judging device is the image in a first variation of the received image of the region exceeding a first variation comparison value. The abnormality determination signal is issued when it is determined that the abnormality and a second degree of fluctuation of the regional area sound signal exceeds a first variation comparison value and the sound is abnormally determined. 如申請專利範圍第1項所述之基於聲音與影像變化的異常判斷系統,其中該異常判斷裝置所接收到的該區域影像的一第一變動程度超過一第一變動比較值、以及該區域區域聲音訊號的一第二變動程度超過一第二變動比較值時,發出該異常判斷訊號。 An abnormality judging system based on sound and image change according to claim 1, wherein a first variation degree of the image of the region received by the abnormality determining device exceeds a first variation comparison value, and the region region When the second variation degree of the sound signal exceeds a second variation comparison value, the abnormality determination signal is issued. 如申請專利範圍第3項所述之基於聲音與影像變化的異常判斷系統,其中該異常判斷裝置根據在一第一時間所接收到的該區域影像、以及在一第二時間所接收到的該區域影像來比較出一差異影像,其中該第一時間早於該第二時間。 An abnormality judging system based on sound and image change as described in claim 3, wherein the abnormality judging means receives the image of the area received at a first time and the second time received The area image compares a difference image, wherein the first time is earlier than the second time. 如申請專利範圍第3項所述之基於聲音與影像變化的異常判斷系統,其中該第一時間所接收到的該區域影像、以及在該第二時間所接收到的該區域影像為對齊,以便補償該攝影裝置在移動攝影模式下所造成的變動。 An abnormality judging system based on sound and image changes as described in claim 3, wherein the image of the area received at the first time and the image of the area received at the second time are aligned so that The variation caused by the photographing device in the mobile photography mode is compensated. 如申請專利範圍第4項所述之基於聲音與影像變化的異常判斷系統,其中該異常判斷裝置在該第一時間與第二時間內所比較出的屬於該第一變動程度的該差異影像超過該第一變動比較值、以及該區域區域聲音訊號的該第二變動程度超過第二變動比較值時,發出該異常判斷訊號。 An abnormality judging system based on sound and image changes as described in claim 4, wherein the difference image that the abnormality judging device compares in the first time and the second time belongs to the first degree of variation exceeds The abnormality determination signal is issued when the first variation comparison value and the second variation degree of the regional region audio signal exceed the second variation comparison value. 如申請專利範圍第3項所述之基於聲音與影像變化的異常 判斷系統,其中該異常判斷裝置根據在一第一時間所接收到的該區域影像、以及在一預訂時間內所接收到的該區域影像來比較出一差異影像。 Anomalies based on sound and image changes as described in item 3 of the patent application scope a judging system, wherein the abnormality determining device compares a difference image according to the image of the area received at a first time and the image of the area received during a predetermined time. 如申請專利範圍第項7所述之基於聲音與影像變化的異常判斷系統,其中該異常判斷裝置在該第一時間、以及該預訂時間內所比較出的屬於該第一變動程度的該差異影像超過該第一變動比較值、以及該區域區域聲音訊號的該第二變動程度超過第二變動比較值時,發出該異常判斷訊號。 The abnormality judging system based on sound and image change according to Item 7 of the patent application, wherein the difference judging device compares the difference image belonging to the first degree of change at the first time and the reservation time. The abnormality determination signal is issued when the first variation comparison value and the second variation degree of the regional region audio signal exceed the second variation comparison value. 如申請專利範圍第4或7項所述之基於聲音與影像變化的異常判斷系統,其中該異常判斷裝置利用景深地圖(depth map)技術判斷出該差異影像中出現異常物的一異常物概率分佈,並依據該異物概率分佈,讓該異常判斷裝置利用背景相減技術從該第二時間所接收到的該區域影像分析出一異常物體。 An abnormality judging system based on sound and image change as described in claim 4 or 7, wherein the abnormality judging device uses a depth map technique to determine an abnormality probability distribution of an abnormal object in the difference image. And according to the foreign matter probability distribution, the abnormality determining device uses the background subtraction technique to analyze an abnormal object from the region image received at the second time. 如申請專利範圍第9項所述之基於聲音與影像變化的異常判斷系統,其中當該區域影像中該異常物體所佔據的範圍超過屬於該第一變動比較值的比例值時,以及該區域區域聲音訊號的該第二變動程度超過該第二變動比較值時,發出該異常判斷訊號。 An abnormality judging system based on sound and image change according to claim 9, wherein when the range occupied by the abnormal object in the region image exceeds a proportional value belonging to the first variation comparison value, and the region region When the second degree of change of the sound signal exceeds the second variation comparison value, the abnormality determination signal is issued. 如申請專利範圍第3項所述之基於聲音與影像變化的異常判斷系統,其中該異常判斷裝置將所接收到的該區域區域聲音訊號處理成一數位訊號後,並基於一分貝對照表而產出一區域分貝值,該分貝對照表包括有複數個數位訊號及匹配該數位訊號的分貝值。 The abnormality judging system based on the sound and image change described in claim 3, wherein the abnormality judging device processes the received regional area sound signal into a digital signal, and outputs the data based on a decibel comparison table. An area decibel value, the decibel comparison table includes a plurality of digital signals and a decibel value matching the digital signal. 如申請專利範圍第3項所述之基於聲音與影像變化的異常判斷系統,其中該異常判斷裝置將所接收到的該區域區域聲音訊號處理成一數位訊號後,並依據一攝影距離,基於一距離分貝對照表而產出一區域分貝值,該距離分貝對照表包括有複數個數位訊號及匹配該數位訊號與該攝影距離的分貝值。 An abnormality judging system based on sound and image change as described in claim 3, wherein the abnormality judging device processes the received regional area sound signal into a digital signal, and based on a photographing distance, based on a distance The decibel table produces a region decibel value, and the distance decibel table includes a plurality of digit signals and a decibel value matching the digit signal and the photographing distance. 如申請專利範圍第7項所述之基於聲音與影像變化的異常判斷系統,其中該異常判斷裝置將所接收到的該區域區域聲音訊號處理成一數位訊號後,並依據一攝影距離,基於一距離分貝對照表而產出一區域分貝值,該距離分貝對照表包括有複數個數位訊號及匹配該數位訊號與該攝影距離的分貝值,其中該攝影距離為景深地圖(depth map)技術所判斷出該異常物體與該攝影裝置之間的距離。 The abnormality judging system based on the sound and image change described in claim 7, wherein the abnormality judging device processes the received regional area sound signal into a digital signal, and based on a photographing distance, based on a distance Decibel comparison table generates a regional decibel value, the distance decibel comparison table includes a plurality of digital signals and a decibel value matching the digital signal and the photographic distance, wherein the photographic distance is determined by a depth map technique The distance between the anomalous object and the photographic device. 如申請專利範圍第11或12項或13項所述之基於聲音與影像變化的異常判斷系統,其中該異常判斷裝置根據在一第一時間所接收到的該區域分貝值、以及在一第二時間所接 收到的該區域分貝值來比較出一差異分貝,其中該第一時間早於該第二時間。 An abnormality judging system based on sound and image changes as described in claim 11 or 12 or 13 wherein the abnormality judging means is based on the decibel value of the region received at a first time, and in a second Time received The received decibel value of the region is compared to a difference decibel, wherein the first time is earlier than the second time. 如申請專利範圍第11或12項或13項所述之基於聲音與影像變化的異常判斷系統,其中該異常判斷裝置根據在一第一時間所接收到的該區域分貝值、以及在一預訂時間內所接收到的該區域分貝值來比較出一差異分貝。 An abnormality judging system based on sound and image changes as described in claim 11 or 12 or 13 wherein the abnormality judging means is based on the decibel value of the area received at a first time, and at a reservation time The decibel value of the region received in the region is compared to a difference decibel. 如申請專利範圍第14或15項所述之基於聲音與影像變化的異常判斷系統,其中當該區域影像的該第一變動程度超過一第一變動比較值時,以及屬於該第二變動程度的該差異分貝的超過第二變動比較值時,發出該異常判斷訊號。 An abnormality determining system based on sound and image changes as described in claim 14 or 15, wherein when the first degree of variation of the image of the area exceeds a first variation comparison value, and the second degree of variation When the difference decibel exceeds the second variation comparison value, the abnormality determination signal is issued. 如申請專利範圍第16項所述之基於聲音與影像變化的異常判斷系統,其中僅選擇實體物可發出的一聲音頻段內的該區域分貝值進行比較出該差異分貝。 The abnormality judging system based on sound and image change as described in claim 16 wherein only the decibel value of the region in a sound frequency band that can be emitted by the physical object is selected to compare the difference decibel. 如申請專利範圍第17項所述之基於聲音與影像變化的異常判斷系統,其中該聲音頻段為20Hz~4KHz。 An abnormality judging system based on sound and image changes as described in claim 17 of the patent application, wherein the sound frequency band is 20 Hz to 4 kHz. 一種基於聲音與影像變化的異常判斷系統,用以監控一場所,該異常判斷系統包含:一攝影裝置,用以對準該場所進行拍攝,並產出一場所影像; 一聲音接收裝置,用以對準該場所進行收音,並產出一場所區域聲音訊號;以及一異常判斷裝置,在所接收到的該場所影像的一第一變動程度超過一第一變動比較值、以及該場所區域聲音訊號的一第二變動程度超過一第二變動比較值時,發出一異常判斷訊號。 An abnormality judging system based on sound and image changes for monitoring a place, the abnormality judging system comprising: a photographing device for photographing the place and generating a scene image; a sound receiving device for aligning the location for sound collection and generating a local area sound signal; and an abnormality determining device, wherein the first variation of the received image of the location exceeds a first variation comparison value And an abnormality determination signal is sent when a second degree of change of the sound signal of the location area exceeds a second variation comparison value.
TW104109216A 2015-03-23 2015-03-23 Abnormality determination system based on sound and image changes TW201635247A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
TW104109216A TW201635247A (en) 2015-03-23 2015-03-23 Abnormality determination system based on sound and image changes

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
TW104109216A TW201635247A (en) 2015-03-23 2015-03-23 Abnormality determination system based on sound and image changes

Publications (1)

Publication Number Publication Date
TW201635247A true TW201635247A (en) 2016-10-01

Family

ID=57847377

Family Applications (1)

Application Number Title Priority Date Filing Date
TW104109216A TW201635247A (en) 2015-03-23 2015-03-23 Abnormality determination system based on sound and image changes

Country Status (1)

Country Link
TW (1) TW201635247A (en)

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN111323785A (en) * 2018-12-13 2020-06-23 青岛海尔多媒体有限公司 Obstacle recognition method and laser television

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN111323785A (en) * 2018-12-13 2020-06-23 青岛海尔多媒体有限公司 Obstacle recognition method and laser television

Similar Documents

Publication Publication Date Title
WO2016150045A1 (en) Housebreaking alarm method and device
KR101990803B1 (en) PROTECTION SYSTEM FOR VULNERABLE CLASSES USING Internet Of Things AND METHOD THEREFOR
CN104135642A (en) Intelligent monitoring method and relevant equipment
CN103198605A (en) Indoor emergent abnormal event alarm system
KR20120050073A (en) System for realtime observing fire using cctv camera, and method for the same
WO2017161880A1 (en) Regional monitoring and alarming system and alarming method
CN110602453B (en) Internet of things big data intelligent video security monitoring system
Ray et al. An intelligent vision system for monitoring security and surveillance of ATM
CN209343418U (en) Public security hearing room monitoring system
TW201801520A (en) Image device and method for detecting a chief
JP5758165B2 (en) Article detection device and stationary person detection device
TW201635247A (en) Abnormality determination system based on sound and image changes
KR101046819B1 (en) Method and system for watching an intrusion by software fence
CN103607577A (en) Aged-care at home and security and protection system based on body motion trail analysis
KR20210061182A (en) Security alarm system that works with environmental sensor based on IP camera
CN107437320A (en) A kind of intelligent free-standing smoke sensitive fire hazard detection alarm and its alarm method
CN104269017B (en) A kind of method for supervising
KR20130068334A (en) Automatic recognition and response system of the armed robbers and the methods of the same
CN109272696A (en) Electronic equipment
KR20150015814A (en) Apparatus, method, and recording medium for emergency alert
CN104200609B (en) A kind of electrical alloy intelligence Factory Building
US11941959B2 (en) Premises monitoring using acoustic models of premises
TW569629B (en) Motion detection method with user-adjustable parameters for better detection accuracy
JP2020167500A (en) Image processing apparatus and image processing program
CN112863102A (en) Method and device for detecting fire hazard