TWI671684B - System and method for monitoring an image - Google Patents
System and method for monitoring an image Download PDFInfo
- Publication number
- TWI671684B TWI671684B TW107145246A TW107145246A TWI671684B TW I671684 B TWI671684 B TW I671684B TW 107145246 A TW107145246 A TW 107145246A TW 107145246 A TW107145246 A TW 107145246A TW I671684 B TWI671684 B TW I671684B
- Authority
- TW
- Taiwan
- Prior art keywords
- identified
- image
- image capturing
- unit
- identification
- Prior art date
Links
Landscapes
- Closed-Circuit Television Systems (AREA)
- Image Analysis (AREA)
Abstract
一種影像監視方法及系統，用以解決習知視訊監控系統影像辨識效率不佳的問題。係包含：控制數個影像擷取單元朝一監視區域內拍攝，以各別產生一監視畫面；當該數個影像擷取單元中的任一影像擷取單元拍攝到一待辨識物時，確認該待辨識物是否已被標記一識別身分於該影像擷取單元的監視畫面上，若確認為否，透過該影像擷取單元的交集視野關聯表進行判斷，判斷與該影像擷取單元具有關聯性的另一影像擷取單元，是否同時拍攝到該待辨識物，且已知該待辨識物的識別身分，若判斷為是，將該待辨識物的識別身分儲存於該影像擷取單元的交集視野關聯表。An image monitoring method and system are used to solve the problem of poor image recognition efficiency of a conventional video monitoring system. The system includes: controlling a plurality of image capturing units to shoot in a surveillance area to generate a monitoring screen respectively; when any one of the image capturing units in the plurality of image capturing units captures an object to be identified, confirm the Whether the object to be identified has been marked with an identification identity on the monitoring screen of the image capture unit, if it is not determined, it is judged through the intersection field correlation table of the image capture unit, and it is determined to be related to the image capture unit Another image capture unit of the camera, whether the to-be-identified object is captured at the same time, and the identification identity of the to-be-identified object is known, and if it is determined to be yes, the identification identity of the to-be-identified object is stored at the intersection of the image capture unit View association table.
Description
本發明係關於一種影像監視方法及系統，尤其是一種能夠讓數個監控攝影機彼此之間互相分享辨識結果的影像監視方法及系統。The present invention relates to an image monitoring method and system, and more particularly to an image monitoring method and system capable of allowing several surveillance cameras to share identification results with each other.
習知影像監控系統，在運用多支監控攝影機對一封閉區域進行安全監控時，由於無法整合每部監控攝影機所拍攝到的物件的資訊，因此，當每一部監控攝影機拍攝到一個未知的待辨識物時，都需重新辨識而導致後端系統的負荷而造成影像辨識效率下降。Known image surveillance systems. When multiple surveillance cameras are used to securely monitor a closed area, the information of objects captured by each surveillance camera cannot be integrated. Therefore, when each surveillance camera captures an unknown When identifying objects, they need to be re-identified, which results in a load on the back-end system and reduces the image recognition efficiency.
有鑑於此，習知的影像監控系統確實仍有加以改善之必要。In view of this, the conventional video surveillance system does still need to be improved.
為解決上述問題，本發明的目的是提供一種影像監視方法，能夠讓數個監控攝影機彼此之間互相分享辨識結果。In order to solve the above problems, an object of the present invention is to provide an image monitoring method, which can allow several surveillance cameras to share identification results with each other.
本發明的次一目的是提供一種影像監視系統，能夠讓數個監控攝影機彼此之間互相分享辨識結果。A secondary object of the present invention is to provide an image monitoring system, which can allow several surveillance cameras to share the identification results with each other.
本發明全文所述之「關聯性」，係指架設於同一個監視區域內的任意二影像擷取單元，分別對該監視區域內的同一個場景進行拍攝，以各別產生一監視畫面，當該二監視畫面之間具有相同的特徵點時，則該二影像擷取單元之間係具有一關聯性。The “relevance” described in the full text of the present invention refers to any two image capturing units set up in the same surveillance area, respectively, to shoot the same scene in the surveillance area to generate a surveillance picture respectively. When the two monitoring pictures have the same feature points, there is a correlation between the two image capturing units.
本發明全文所述之「光學中心」，係指當來自各方向的光線通過透鏡上的一點時，各光線的傳播方向不會產生變化（如，折射），則該點即為一光學中心，係本發明所屬技術領域中具有通常知識者可以理解。The “optical center” mentioned in the present invention means that when light from each direction passes through a point on the lens, the propagation direction of each light will not change (eg, refraction), then the point is an optical center. It can be understood by those having ordinary knowledge in the technical field to which the present invention belongs.
本發明的影像監視方法，包含：建立一交集視野關聯表，該交集視野關聯表係用以記錄數個影像擷取單元彼此間的關聯性，該數個影像擷取單元係分別架設於同一個監視區域內；控制該數個影像擷取單元朝該監視區域內拍攝，以各別產生一監視畫面；及當該數個影像擷取單元中的任一影像擷取單元拍攝到一待辨識物時，確認該待辨識物是否已被標記一識別身分於該影像擷取單元的監視畫面上，若確認結果為否，則透過該影像擷取單元的交集視野關聯表進行判斷，判斷與該影像擷取單元具有關聯性的另一影像擷取單元，是否同時拍攝到該待辨識物，且已知該待辨識物的識別身分，若判斷結果為是，則將該待辨識物的識別身分儲存於該影像擷取單元的交集視野關聯表。The image monitoring method of the present invention includes: establishing an intersection view association table, the intersection view association table is used to record the correlation between a plurality of image capturing units, and the plurality of image capturing units are respectively set up on the same Within the monitoring area; controlling the plurality of image capturing units to shoot toward the monitoring area to generate a monitoring screen respectively; and when any one of the plurality of image capturing units captures an object to be identified When confirming whether the object to be identified has been marked with an identification identity on the monitoring screen of the image capturing unit, and if the result of the confirmation is no, then judging through the intersection visual field association table of the image capturing unit, and determining the relationship with the image Whether the capture unit has another image capture unit associated with it, whether the object to be identified is captured at the same time, and the identification identity of the object to be identified is known, and if the determination result is yes, the identification identity of the object to be identified is stored Intersection view association table at the image capturing unit.
本發明的影像監視系統，包含：數個影像擷取單元，分別架設於同一個監視區域內；一儲存單元，用以儲存各該影像擷取單元的一交集視野關聯表；及一處理單元，耦接該數個影像擷取單元及該儲存單元，該處理單元分析該數個影像擷取單元彼此間的關聯性，以分別產生各該影像擷取單元的交集視野關聯表，控制該數個影像擷取單元朝該監視區域內拍攝，以各別產生一監視畫面，當該數個影像擷取單元中的任一影像擷取單元拍攝到一待辨識物時，該處理單元確認該待辨識物是否已被標記一識別身分於該影像擷取單元的監視畫面上，若確認結果為否，則透過該影像擷取單元的交集視野關聯表進行判斷，判斷與該影像擷取單元具有關聯性的另一影像擷取單元是否同時拍攝到該待辨識物，且已知該辨識物的識別身分，若判斷結果為是，則將該待辨識物的識別身分儲存於該影像擷取單元的交集視野關聯表。The image monitoring system of the present invention comprises: a plurality of image capturing units respectively set up in the same monitoring area; a storage unit for storing an intersection view association table of each of the image capturing units; and a processing unit, The image capturing units and the storage unit are coupled, and the processing unit analyzes the correlation between the image capturing units to generate an intersection view association table for each of the image capturing units and controls the plurality of image capturing units. The image capturing unit shoots towards the monitoring area to generate a monitoring screen respectively. When any image capturing unit of the plurality of image capturing units captures an object to be identified, the processing unit confirms the object to be identified Whether the object has been marked with an identification identity on the monitoring screen of the image capturing unit. If the result of the confirmation is no, then it is judged through the intersection view association table of the image capturing unit, and it is determined to be related to the image capturing unit. Whether another image capture unit of the camera simultaneously captures the to-be-identified object, and the identification identity of the to-be-identified object is known, and if the determination result is yes, the Do identity association table stored in the intersection of the field of view of the image capturing unit.
據此，本發明的影像監視方法及系統，係能夠藉由該交集視野關聯表得知各該影像擷取單元之間的關聯性，並透過分析一影像擷取單元所拍攝到的一待辨識物，是否已被具有關聯性的另一影像擷取單元得知該待辨識物的識別身分，以分享該待辨識物的識別身分給該影像擷取單元。如此，本發明的影像監視方法及系統能夠降低後端系統的運算負荷，以便將更多的系統資源用於辨識影像上，係具有提升影像辨識效率的功效。According to this, the image monitoring method and system of the present invention can learn the correlation between the image capturing units through the intersection field of view association table, and analyze an image to be identified captured by an image capturing unit. Whether the object has been identified by another image capture unit having an association with the identification identity of the object to be identified, so as to share the identification identity of the object to be identified with the image capture unit. In this way, the image monitoring method and system of the present invention can reduce the computing load of the back-end system, so that more system resources are used for identifying the image, which has the effect of improving the efficiency of image identification.
其中，若該待辨識物已被標記該識別身分於該影像擷取單元所拍攝的監視畫面上，則儲存一影像位置座標於該影像擷取單元的交集視野關聯表，該影像位置座標係為該待辨識物於該監視畫面上的位置座標。如此，係具有紀錄該待辨識物於該監視區域內的移動路徑的功效。Wherein, if the to-be-identified object has been marked with the identification identity on the monitoring screen captured by the image capturing unit, an image position coordinate is stored in an intersection view association table of the image capturing unit, and the image position coordinate is The position coordinates of the object to be identified on the monitoring screen. In this way, it has the effect of recording the movement path of the object to be identified in the surveillance area.
其中，當符合下列公式時，則該影像擷取單元與該另一影像擷取單元係同時拍到該待辨識物， ，其中，F係表示為基本矩陣，P A，P B係分別表示為該影像擷取單元與該另一影像擷取單元拍攝到該待辨識物時，於各自的一成像平面上所產生的一投影點，該二成像平面分別具有一光學中心，該二光學中心與該待辨識物形成一共平面，且該二光學中心相連接以形成一基準線，二垂直線分別通過該二投影點，且分別延伸相交於該基準線上，以各自形成一相交點。如此，係能夠減少系統資源的消耗，係具有提升系統運作效率的功效。 When the following formula is met, the image capturing unit and the other image capturing unit simultaneously capture the object to be identified. Among them, F is represented as a basic matrix, and P A and P B are respectively represented as the images generated by the image capturing unit and the other image capturing unit on the respective imaging plane when the object to be identified is captured. A projection point, the two imaging planes each have an optical center, the two optical centers and the object to be identified form a coplanar plane, and the two optical centers are connected to form a reference line, and two vertical lines pass through the two projection points, The intersections are respectively extended on the reference line to form an intersection point. In this way, the system can reduce the consumption of system resources, and has the effect of improving the efficiency of system operation.
其中，當無法由該另一影像擷取單元取得該待辨識物的識別身分時，控制該影像擷取單元的一影像辨識模型對該待辨識物進行辨識，以取得該待辨識物的識別身分，並將該待辨識物的識別身分儲存於該影像擷取單元的交集視野關聯表。如此，係具有影像辨識的功效。Wherein, when the identification identity of the object to be identified cannot be obtained by the other image capture unit, an image recognition model controlling the image capture unit is used to identify the identification object to obtain the identification identity of the object to be identified. And storing the identification identity of the object to be identified in an intersection view association table of the image capturing unit. In this way, it has the function of image recognition.
其中，當該另一影像擷取單元同時拍攝到該待辨識物，且已知該待辨識物的識別身分時，若該影像擷取單元之交集視野關聯表中，係具有一辨識失敗的待辨識物之影像時，則控制該影像擷取單元的影像辨識模型，依據由該另一影像擷取單元所取得的待辨識物之識別身分重新進行比對，以比對出該辨識失敗的待辨識物的識別身分。如此，該影像辨識模型一開始不需採集太多該待辨識物的影像樣本，可以透過後期的訓練機制逐漸提升該影像辨識模型的辨識率，係具有提升影像辨識模型的辨識能力的功效。Wherein, when the other image capturing unit captures the object to be identified at the same time, and the identification identity of the object to be identified is known, if the intersection field of vision association table of the image capturing unit has an identification failure, When the image of the object is identified, the image recognition model of the image capturing unit is controlled, and the comparison is performed again according to the identification identity of the object to be identified obtained by the other image capturing unit, so as to compare the waiting of the identification failure. The identity of the identifier. In this way, the image recognition model does not need to collect too many image samples of the object to be identified at first, and can gradually improve the recognition rate of the image recognition model through a later training mechanism, which has the effect of improving the recognition ability of the image recognition model.
其中，該處理單元確認該影像擷取單元所拍攝的監視畫面上，該待辨識物是否已被標記該識別身分，若確認結果為是，則儲存一影像位置座標於該影像擷取單元的交集視野關聯表，該影像位置座標係為該待辨識物於該監視畫面上的位置座標。如此，係具有紀錄該待辨識物於該監視區域內的移動路徑的功效。Wherein, the processing unit confirms whether the object to be identified is marked with the identification on the monitoring screen captured by the image capturing unit, and if the confirmation result is yes, stores an image position coordinate at the intersection of the image capturing unit In the visual field correlation table, the image position coordinates are the position coordinates of the object to be identified on the monitoring screen. In this way, it has the effect of recording the movement path of the object to be identified in the surveillance area.
其中，當符合下列公式時，則該影像擷取單元與該另一影像擷取單元係同時拍到該待辨識物， ，其中，F係表示為基本矩陣，P A，P B係分別表示為該影像擷取單元與該另一影像擷取單元拍攝到該待辨識物時，於各自的一成像平面上所產生的一投影點，該二成像平面分別具有一光學中心，該二光學中心與該待辨識物形成一共平面，且該二光學中心相連接以形成一基準線，二垂直線分別通過該二投影點，且分別延伸相交於該基準線上，以各自形成一相交點。如此，能夠減少系統資源的消耗，係具有提升系統運作效率的功效。 When the following formula is met, the image capturing unit and the other image capturing unit simultaneously capture the object to be identified. Among them, F is represented as a basic matrix, and P A and P B are respectively represented as the images generated by the image capturing unit and the other image capturing unit on the respective imaging plane when the object to be identified is captured. A projection point, the two imaging planes each have an optical center, the two optical centers and the object to be identified form a coplanar plane, and the two optical centers are connected to form a reference line, and two vertical lines pass through the two projection points, The intersections are respectively extended on the reference line to form an intersection point. In this way, the consumption of system resources can be reduced, which has the effect of improving the system operation efficiency.
本發明的影像監視系統還可以另包含數個影像辨識模型，該數個影像辨識模型耦接該處理單元，各該影像辨識模型用以辨識該待辨識物，以取得該待辨識物的識別身分，當無法由該另一影像擷取單元取得該待辨識物的識別身分時，該處理單元控制該影像擷取單元的影像辨識模型對該待辨識物進行辨識，以取得該待辨識物的識別身分，並將該待辨識物的識別身分儲存於該影像擷取單元的交集視野關聯表。如此，係具有影像辨識的功效。The image monitoring system of the present invention may further include a plurality of image recognition models, the plurality of image recognition models are coupled to the processing unit, and each of the image recognition models is used to identify the object to be identified, so as to obtain the identification identity of the object to be identified. , When the identification identity of the to-be-identified object cannot be obtained by the other image capturing unit, the processing unit controls the image recognition model of the image capturing unit to identify the to-be-identified object to obtain the identification of the to-be-identified object Identity, and the identity of the object to be identified is stored in an intersection view association table of the image capturing unit. In this way, it has the function of image recognition.
其中，當該另一影像擷取單元同時拍攝到該待辨識物，且已知該辨識物的識別身分時，若該影像擷取單元之交集視野關聯表中，係具有一辨識失敗的待辨識物之影像時，則該處理單元控制該影像擷取單元的影像辨識模型，依據由該另一影像擷取單元所取得的待辨識物之識別身分重新進行比對，以比對出該辨識失敗的待辨識物的識別身分。如此，該影像辨識模型一開始不需採集太多該待辨識物的影像樣本，可以透過後期的訓練機制逐漸提升該影像辨識模型的辨識率，係具有提升影像辨識模型的辨識能力的功效。Wherein, when the other image capturing unit captures the object to be identified at the same time, and the identification identity of the object is known, if the intersection view association table of the image capturing unit has an identification failure to be identified When the object is imaged, the processing unit controls the image recognition model of the image capture unit, and re-compares according to the identification identity of the object to be identified obtained by the other image capture unit to compare the recognition failure. Identification of the object to be identified. In this way, the image recognition model does not need to collect too many image samples of the object to be identified at first, and can gradually improve the recognition rate of the image recognition model through a later training mechanism, which has the effect of improving the recognition ability of the image recognition model.
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：In order to make the above and other objects, features, and advantages of the present invention more comprehensible, the following describes the preferred embodiments of the present invention and the accompanying drawings in detail, as follows:
請參照第1圖所示，其係本發明影像監視方法的一較佳實施例，係包含一建立步驟S1、一偵測步驟S2、一搜尋步驟S3及一更新步驟S4。Please refer to FIG. 1, which is a preferred embodiment of the image monitoring method of the present invention, which includes a establishing step S1, a detecting step S2, a searching step S3 and an updating step S4.
該建立步驟S1係用以建立一交集視野關聯表，該交集視野關聯表係用以紀錄數個影像擷取單元（如：監視器）彼此間的關聯性，該數個影像擷取單元係分別架設於同一個監視區域內。當該數個影像擷取單元中的任意二影像擷取單元係具有關聯性時，將該二影像擷取單元的關聯性各別記錄於各自的交集視野關聯表。在本實施例中，該交集視野關聯表係可以包含一影像擷取單元編號、一識別身分、一關聯擷取單元編號、一座標及一基本矩陣，該交集視野關聯表可如下表一所示。較佳地，還可以具有一更新時間及一辨識失敗的待辨識物之影像。The establishing step S1 is used to establish an intersection view association table. The intersection view association table is used to record the correlation between a plurality of image capture units (such as monitors). The plurality of image capture units are respectively Erected in the same surveillance area. When any two image capturing units of the plurality of image capturing units are related, the correlations of the two image capturing units are recorded in their respective intersection view association tables. In this embodiment, the intersection view association table may include an image capture unit number, an identification identity, a correlation capture unit number, a target, and a basic matrix. The intersection view association table may be shown in Table 1 below. . Preferably, it may also have an update time and an image of the object to be identified that fails to be identified.
表一 交集視野關聯表
該偵測步驟S2係以一臉部偵測軟體（如：Yolo）偵測上述影像擷取單元所拍攝的監視畫面中，是否具有一待辨識物（如：人臉影像）。該臉部偵測軟體中係可以內建有各該待辨識物的一影像，以及相對應的識別身分。當該監視畫面中不具有該待辨識物時，則持續進行偵測；反之，當該數個影像擷取單元中的任一影像擷取單元拍攝到一待辨識物時，則進一步確認該待辨識物是否已被標記一識別身分於該影像擷取單元的監視畫面上，若確認結果為否，則執行該搜尋步驟S3；若確認結果為是，則取得該待辨識物於該影像擷取單元之監視畫面中的影像位置座標，並執行該更新步驟S4。The detecting step S2 is to use a face detection software (such as Yolo) to detect whether there is an object to be identified (such as a face image) in the monitoring picture captured by the image capturing unit. The face detection software may have an image of each of the objects to be identified, and a corresponding identification identity. When the object to be identified is not included in the monitoring screen, detection is continuously performed; otherwise, when any one of the image capturing units captures an object to be identified, the object to be identified is further confirmed. Whether the identification object has been marked with an identification identity on the monitoring screen of the image capture unit, if the confirmation result is no, then execute the search step S3; if the confirmation result is yes, then obtain the object to be identified in the image capture The coordinates of the image position in the monitoring screen of the unit, and the update step S4 is executed.
該搜尋步驟S3係透過該影像擷取單元的交集視野關聯表進行判斷，判斷與該影像擷取單元具有關聯性的另一影像擷取單元，是否同時拍攝到該待辨識物，且該另一影像擷取單元已知該待辨識物的識別身分。若判斷結果為是，則執行該更新步驟S4。The searching step S3 is performed by judging through an intersection view association table of the image capturing unit, and determining whether another image capturing unit that is related to the image capturing unit has simultaneously captured the object to be identified, and the other The image capture unit knows the identification identity of the object to be identified. If the determination result is YES, the update step S4 is executed.
請參照第2圖所示，詳言之，當符合下列公式（1）時，則該影像擷取單元與該另一影像擷取單元係同時拍到該待辨識物。 （1） 其中，F係表示為基本矩陣，P A，P B係分別表示為該影像擷取單元與該另一影像擷取單元拍攝到該待辨識物R時，於各自的一成像平面I A，I B上所產生的一投影點，該二成像平面I A，I B分別具有一光學中心O A，O B，該二光學中心O A，O B與該待辨識物R形成一共平面（coplanar），且該二光學中心O A，O B相連接以形成一基準線L，二垂直線分別通過該二投影點P A，P B，且分別延伸相交於該基準線L上，以各自形成一相交點C A，C B。 Please refer to FIG. 2. In detail, when the following formula (1) is met, the image capturing unit and the other image capturing unit simultaneously capture the object to be identified. (1) Among them, F is represented as the basic matrix, and P A and P B are respectively represented as the image capturing unit and the other image capturing unit when the object R to be identified is captured, in each of the imaging planes I A projection point generated on A , I B , the two imaging planes I A , I B respectively have an optical center O A , O B , and the two optical centers O A , O B and the object R form a common plane (Coplanar), and the two optical centers O A , O B are connected to form a reference line L, and the two vertical lines pass through the two projection points P A , P B , respectively, and extend to intersect on the reference line L, respectively, to Each forms an intersection point C A , C B.
該更新步驟S4係用以更新各該影像擷取單元的交集視野關聯表。在本實施例中，當該待辨識物已被標記該識別身分於該影像擷取單元所拍攝的監視畫面上，則儲存該待辨識物的影像位置座標於該影像擷取單元的交集視野關聯表。又或者，當該另一影像擷取單元與該影像擷取單元是同時拍攝到該待辨識物，且已知該待辨識物的識別身分時，則儲存該待辨識物的識別身分於該影像擷取單元的交集視野關聯表。較佳地，還可以一併儲存該更新步驟S4的一執行時間。The updating step S4 is used to update an intersection view association table of each image capturing unit. In this embodiment, when the to-be-identified object has been marked with the identification identity on the monitoring screen captured by the image capturing unit, the image position coordinates of the to-be-identified object are stored in the intersection field of view association of the image capturing unit. table. Or, when the other image capturing unit and the image capturing unit capture the object to be identified at the same time, and the identification identity of the object to be identified is known, the identification identity of the object to be identified is stored in the image. Intersection view association table of extraction units. Preferably, an execution time of the update step S4 can also be stored together.
較佳地，當無法由該另一影像擷取單元取得該待辨識物的識別身分時，即該搜尋步驟S3的判斷結果為否時，本發明的影像監視方法係還可以另包含一辨識步驟S5，該辨識步驟S5係用以控制該影像擷取單元的臉部偵測軟體的一影像辨識模型對該待辨識物進行辨識，以取得該待辨識物的識別身分，並執行該更新步驟S4，以將該待辨識物的識別身分儲存於該影像擷取單元的交集視野關聯表。Preferably, when the identification identity of the object to be identified cannot be obtained by the other image capturing unit, that is, when the determination result of the searching step S3 is no, the image monitoring method of the present invention may further include an identification step. S5. The identification step S5 is used to control an image recognition model of the face detection software of the image capture unit to identify the object to be identified, to obtain the identification identity of the object to be identified, and execute the update step S4. To store the identification identity of the object to be identified in the intersection view association table of the image capturing unit.
較佳地，本發明影像監視方法還可以另包含一後置訓練步驟S6，當該搜尋步驟S3的判斷結果為是時，可以一併執行該後置訓練步驟S6。詳言之，當該影像擷取單元之交集視野關聯表中，係具有一辨識失敗的待辨識物之影像時，該後置訓練步驟S6係可以控制該影像擷取單元的影像辨識模型，依據由該另一影像擷取單元所取得的待辨識物之識別身分重新進行比對，以比對出該辨識失敗的待辨識物的識別身分。Preferably, the image surveillance method of the present invention may further include a post-training step S6, and when the determination result of the search step S3 is yes, the post-training step S6 may be performed together. In detail, when the intersection field-of-view correlation table of the image capturing unit is an image of an object to be identified that fails to be identified, the post-training step S6 can control the image recognition model of the image capturing unit, according to The identification identities of the objects to be identified obtained by the another image capturing unit are compared again to compare the identification identities of the objects to be identified that fail to be identified.
請參照第3圖所示，其係本發明影像監視系統的一較佳實施例，係包含數個影像擷取單元1、一儲存單元2及一處理單元3，該處理單元3耦接該數個影像擷取單元1及該儲存單元2。Please refer to FIG. 3, which is a preferred embodiment of the image monitoring system of the present invention, which includes a plurality of image capturing units 1, a storage unit 2 and a processing unit 3. The processing unit 3 is coupled to the data Image capture unit 1 and the storage unit 2.
該數個影像擷取單元1係分別架設於同一個監視區域內，並用以朝該監視區域中的各個區域分別進行拍攝，以各別產生一監視畫面。較佳地，該數個影像擷取單元1所拍攝的監視畫面進行組合拼湊後，係可以覆蓋整個該監視區域。例如但不限制地，各該影像擷取單元1係可以為一相機或一紅外線攝影機。The plurality of image capturing units 1 are respectively set up in the same monitoring area, and are used for shooting towards each area in the monitoring area to generate a monitoring picture respectively. Preferably, the surveillance images captured by the plurality of image capturing units 1 can be combined to piece together to cover the entire surveillance area. For example, without limitation, each of the image capturing units 1 may be a camera or an infrared camera.
該儲存單元2係可以為任何用以儲存電子資料的一儲存媒體，例如可以為：一機械式硬碟（HDD）、一固態硬碟（SSD）或一快閃記憶體（Flash Memory）等非揮發性記憶體。該儲存單元2係用以儲存各該影像擷取單元1的一交集視野關聯表。在本實施例中，該交集視野關聯表係可以包含一影像擷取單元編號、一識別身分、一關聯擷取單元編號、一座標及一基本矩陣。較佳地，還可以具有一更新時間及一辨識失敗的待辨識物之影像。The storage unit 2 can be any storage medium used to store electronic data. For example, the storage unit 2 can be a mechanical hard disk (HDD), a solid-state hard disk (SSD), or a flash memory (Flash Memory). Volatile memory. The storage unit 2 is used to store an intersection view association table of each of the image capturing units 1. In this embodiment, the intersection view association table may include an image capture unit number, an identification identity, an associated capture unit number, a target, and a basic matrix. Preferably, it may also have an update time and an image of the object to be identified that fails to be identified.
該處理單元3耦接該數個影像擷取單元1及該儲存單元2，該處理單元3係可以為任何具有資料多工運算、訊號產生及控制等功能的一電子晶片，如：可程式邏輯控制器（PLC）、數位訊號處理器（DSP）、微控制器（MCU）或具上述功能的電路板等。簡而言之，該處理單元3分析該數個影像擷取單元1中的任意二影像擷取單元1所拍攝的監視畫面，是否具有相同的特徵點，若分析結果為是，則將該二影像擷取單元1的關聯性紀錄於各自的交集視野關聯表。再且，當該二影像擷取單元1的其中一影像擷取單元1拍攝到一待辨識物時，該處理單元4可藉由該交集視野關聯表，由該二影像擷取單元1的其中另一影像擷取單元1得知該待辨識物之識別身分。The processing unit 3 is coupled to the image capturing units 1 and the storage unit 2. The processing unit 3 can be any electronic chip with functions such as data multiplexing, signal generation and control, such as programmable logic Controller (PLC), digital signal processor (DSP), microcontroller (MCU) or circuit board with the above functions. In short, the processing unit 3 analyzes whether the monitoring images captured by any two of the image capturing units 1 have the same feature points. If the analysis result is yes, then the two The relevance of the image capturing unit 1 is recorded in the respective intersection view association table. Furthermore, when one of the two image capturing units 1 captures an object to be identified, the processing unit 4 may use the intersection field of view association table to select one of the two image capturing units 1 from among them. Another image capturing unit 1 learns the identification identity of the object to be identified.
具體而言，該處理單元3分析該數個影像擷取單元1彼此間的關聯性，以分別產生各該影像擷取單元1的交集視野關聯表。該處理單元3控制該數個影像擷取單元1朝該監視區域內拍攝，以各別產生一監視畫面。當該數個影像擷取單元1中的任一影像擷取單元1拍攝到一待辨識物時，該處理單元3確認該待辨識物是否已被標記一識別身分於該影像擷取單元1的監視畫面上，若確認結果為是，則儲存一影像位置座標於該影像擷取單元1的交集視野關聯表，該影像位置座標係為該待辨識物於該監視畫面上的位置座標；若確認結果為否，則透過該影像擷取單元的交集視野關聯表進行判斷，判斷與該影像擷取單元1具有關聯性的另一影像擷取單元1是否同時拍攝到該待辨識物，且該另一影像擷取單元1已知該辨識物的識別身分，若判斷結果為是，即符合上述公式（1），則將該待辨識物的識別身分儲存於該影像擷取單元1的交集視野關聯表。Specifically, the processing unit 3 analyzes the correlation between the image capturing units 1 to generate an intersection view association table of the image capturing units 1. The processing unit 3 controls the plurality of image capturing units 1 to shoot into the monitoring area to generate a monitoring picture respectively. When any one of the image capturing units 1 captures an object to be identified, the processing unit 3 confirms whether the object to be identified has been marked with an identifying identity in the image capturing unit 1 On the monitoring screen, if the confirmation result is yes, an image position coordinate is stored in the intersection view association table of the image capturing unit 1, and the image position coordinate is the position coordinate of the object to be identified on the monitoring screen; if confirmed If the result is no, it is judged through the intersection view association table of the image capturing unit to determine whether another image capturing unit 1 that is related to the image capturing unit 1 simultaneously captures the object to be identified, and the other An image capture unit 1 knows the identification identity of the identified object, and if the determination result is yes, that is, the formula (1) is met, then the identification identity of the object to be identified is stored in the intersection view association of the image capture unit 1 table.
較佳地，本發明的影像監視系統還可以另包含數個影像辨識模型4，各該影像辨識模型4耦接該處理單元3，並用以辨識該待辨識物以取得該待辨識物的識別身分。各該影像辨識模型4係可以基於卷積神經網路（CNN）設計訓練完成，例如可以為YOLO。該卷積神經網路的技術係本發明所屬技術領域中具有通常知識者可以理解，在此不多加贅述。承上述，若判斷結果為否，即無法由該另一影像擷取單元1取得該待辨識物的識別身分時，該處理單元3可控制該影像擷取單元1的影像辨識模型4對該待辨識物進行辨識，以取得該待辨識物的識別身分，並將該待辨識物的識別身分儲存於該影像擷取單元1的交集視野關聯表。Preferably, the image monitoring system of the present invention may further include a plurality of image recognition models 4, each of which is coupled to the processing unit 3 and used to identify the object to be identified to obtain the identification identity of the object to be identified. . Each of the 4 image recognition models can be designed and trained based on a Convolutional Neural Network (CNN), for example, it can be YOLO. The technology of the convolutional neural network can be understood by those having ordinary knowledge in the technical field to which the present invention belongs, and it will not be described in detail here. Following the above, if the determination result is no, that is, when the identification identity of the object to be identified cannot be obtained by the other image capture unit 1, the processing unit 3 may control the image recognition model 4 of the image capture unit 1 to the target The identification object performs identification to obtain the identification identity of the object to be identified, and stores the identification identity of the object to be identified in the intersection view association table of the image capturing unit 1.
承上述，若判斷結果為是，即該另一影像擷取單元1已知該辨識物的識別身分時，則當該影像擷取單元1之交集視野關聯表中，係具有一辨識失敗的待辨識物之影像時，該處理單元3係可以控制該影像擷取單元1的影像辨識模型4，依據由該另一影像擷取單元1所取得的待辨識物之識別身分重新進行比對，以比對出該辨識失敗的待辨識物的識別身分。According to the above, if the determination result is yes, that is, when the other image capturing unit 1 knows the identification identity of the identifier, when the intersection field of vision association table of the image capturing unit 1 has a pending recognition failure, When recognizing the image of the object, the processing unit 3 can control the image recognition model 4 of the image capture unit 1 and re-compare according to the recognition identity of the object to be recognized obtained by the other image capture unit 1 to Compare the identification identities of the objects to be identified that fail to be identified.
綜上所述，本發明的影像監視方法及系統，係能夠藉由該交集視野關聯表得知各該影像擷取單元之間的關聯性，並透過分析一影像擷取單元所拍攝到的一待辨識物，是否已被具有關聯性的另一影像擷取單元得知該待辨識物的識別身分，以分享該待辨識物的識別身分給該影像擷取單元。如此，本發明的影像監視方法及系統能夠降低後端系統的運算負荷，以便將更多的系統資源用於辨識影像上，係具有提升影像辨識效率的功效。In summary, the image monitoring method and system of the present invention can learn the correlation between the image capturing units by using the intersection view association table, and analyze an image captured by an image capturing unit. Whether the to-be-identified object has been recognized by another image capture unit having an association with the identity of the to-be-identified object, and to share the identity of the to-be-identified object to the image capture unit. In this way, the image monitoring method and system of the present invention can reduce the computing load of the back-end system, so that more system resources are used for identifying the image, which has the effect of improving the efficiency of image identification.
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。Although the present invention has been disclosed using the above-mentioned preferred embodiments, it is not intended to limit the present invention. Anyone skilled in the art can make various changes and modifications to the above embodiments without departing from the spirit and scope of the present invention. The technical scope protected by the invention, so the scope of protection of the present invention shall be determined by the scope of the appended patent application.
﹝本發明﹞﹝this invention﹞
S1‧‧‧建立步驟 S1‧‧‧ establishment steps
S2‧‧‧偵測步驟 S2‧‧‧ Detection steps
S3‧‧‧搜尋步驟 S3‧‧‧Search steps
S4‧‧‧更新步驟 S4‧‧‧ update steps
S5‧‧‧辨識步驟 S5‧‧‧Identification steps
S6‧‧‧後置訓練步驟 S6‧‧‧ post training steps
1‧‧‧影像擷取單元 1‧‧‧Image capture unit
2‧‧‧儲存單元 2‧‧‧Storage unit
3‧‧‧處理單元 3‧‧‧ processing unit
4‧‧‧影像辨識模型 4‧‧‧Image recognition model
CA，CB‧‧‧相交點C A , C B ‧‧‧ intersection
IA，IB‧‧‧成像平面I A , I B ‧‧‧ imaging plane
L‧‧‧基準線 L‧‧‧ baseline
OA，OB‧‧‧光學中心O A , O B ‧‧‧ Optical Center
PA，PB‧‧‧投影點P A , P B ‧‧‧ Projection point
R‧‧‧待辨識物 R‧‧‧ To be identified
［第1圖］ 本發明一較佳實施例的方法流程圖。 ［第2圖］ 本發明一較佳實施例用以判斷二影像擷取單元是否拍攝到同一個待辨識物的幾何關係圖。 ［第3圖］ 本發明一較佳實施例的系統方塊圖。[Figure 1] A method flowchart of a preferred embodiment of the present invention. [Figure 2] A preferred embodiment of the present invention is used to determine whether the two image capturing units have captured the geometric relationship diagram of the same object to be identified. [Figure 3] A system block diagram of a preferred embodiment of the present invention.
Claims (10)
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| TW107145246A TWI671684B (en) | 2018-12-14 | 2018-12-14 | System and method for monitoring an image |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| TW107145246A TWI671684B (en) | 2018-12-14 | 2018-12-14 | System and method for monitoring an image |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| TWI671684B true TWI671684B (en) | 2019-09-11 |
| TW202022690A TW202022690A (en) | 2020-06-16 |
Family
ID=68618930
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| TW107145246A TWI671684B (en) | 2018-12-14 | 2018-12-14 | System and method for monitoring an image |
Country Status (1)
| Country | Link |
|---|---|
| TW (1) | TWI671684B (en) |
Families Citing this family (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| TWI774163B (en) * | 2020-12-23 | 2022-08-11 | 國立屏東科技大學 | Radiographic equipment and method of operation |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| TW201225631A (en) * | 2010-09-30 | 2012-06-16 | Apple Inc | Image signal processor line buffer configuration for processing raw image data |
| US8937651B2 (en) * | 2007-04-19 | 2015-01-20 | Dvp Technologies Ltd. | Imaging system and method for use in monitoring a field of regard |
| TW201614997A (en) * | 2010-09-30 | 2016-04-16 | Apple Inc | System, method and electronic device for processing an image signal |
| TW201732556A (en) * | 2015-12-23 | 2017-09-16 | 英特爾股份有限公司 | Hardware content-associative data structure for acceleration of set operations |
-
2018
- 2018-12-14 TW TW107145246A patent/TWI671684B/en active
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US8937651B2 (en) * | 2007-04-19 | 2015-01-20 | Dvp Technologies Ltd. | Imaging system and method for use in monitoring a field of regard |
| TW201225631A (en) * | 2010-09-30 | 2012-06-16 | Apple Inc | Image signal processor line buffer configuration for processing raw image data |
| TW201614997A (en) * | 2010-09-30 | 2016-04-16 | Apple Inc | System, method and electronic device for processing an image signal |
| TW201732556A (en) * | 2015-12-23 | 2017-09-16 | 英特爾股份有限公司 | Hardware content-associative data structure for acceleration of set operations |
Also Published As
| Publication number | Publication date |
|---|---|
| TW202022690A (en) | 2020-06-16 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Lukezic et al. | Cdtb: A color and depth visual object tracking dataset and benchmark | |
| WO2017096949A1 (en) | Method, control device, and system for tracking and photographing target | |
| CN109313799B (en) | Image processing method and apparatus | |
| US9361512B2 (en) | Identification of a gesture | |
| JP5936697B2 (en) | Method, apparatus, and computer-readable recording medium for managing a reference face database to improve face recognition performance in a memory limited environment | |
| CN102833478B (en) | Fault-tolerant background model | |
| US8989506B1 (en) | Incremental image processing pipeline for matching multiple photos based on image overlap | |
| US20130177293A1 (en) | Method and apparatus for the assignment of roles for image capturing devices | |
| CN108198199B (en) | Moving object tracking method, moving object tracking device and electronic equipment | |
| CN110555377B (en) | Pedestrian detection and tracking method based on fish eye camera overlooking shooting | |
| JP6551226B2 (en) | INFORMATION PROCESSING SYSTEM, INFORMATION PROCESSING METHOD, AND PROGRAM | |
| CN101877764A (en) | Camera system and method for carrying out assisted drawing by utilizing same | |
| US10666858B2 (en) | Deep-learning-based system to assist camera autofocus | |
| JP2017228082A (en) | Tracking device, tracking method, and program | |
| CN107770486B (en) | Event search apparatus and system | |
| EP3035242B1 (en) | Method and electronic device for object tracking in a light-field capture | |
| TWI671684B (en) | System and method for monitoring an image | |
| US20190027004A1 (en) | Method for performing multi-camera automatic patrol control with aid of statistics data in a surveillance system, and associated apparatus | |
| González et al. | Single object long-term tracker for smart control of a ptz camera | |
| TWI657378B (en) | Method and system for object tracking in multiple non-linear distortion lenses | |
| JP2007179224A (en) | Information processing device, method, and program | |
| CN114222065B (en) | Image processing method, image processing apparatus, electronic device, storage medium, and program product | |
| US11575826B2 (en) | Image processing apparatus and method, and image capturing apparatus | |
| JP6301202B2 (en) | Shooting condition setting device and shooting condition setting method | |
| TWI690207B (en) | Method for object tracking |