TWI764425B - Real time pedestrian statistical method based on face identification, and apparatus thereof - Google Patents
Real time pedestrian statistical method based on face identification, and apparatus thereofInfo
- Publication number
- TWI764425B TWI764425B TW109143733A TW109143733A TWI764425B TW I764425 B TWI764425 B TW I764425B TW 109143733 A TW109143733 A TW 109143733A TW 109143733 A TW109143733 A TW 109143733A TW I764425 B TWI764425 B TW I764425B
- Authority
- TW
- Taiwan
- Prior art keywords
- face
- pedestrian
- detection
- image
- detection frame
- Prior art date
Links
Images
Landscapes
- Image Analysis (AREA)
- Traffic Control Systems (AREA)
Abstract
Description
本發明涉及一種基於人臉識別的即時行人統計方法以及裝置。 The invention relates to a real-time pedestrian statistical method and device based on face recognition.
隨著科技的發展,人臉識別技術被廣泛的應用在不同領域,例如,在公司內作為通勤刷卡,在通訊領域作為移動設備的解鎖密碼等。基於人臉識別技術,可對圖像內的人流量進行統計。在人群密集程度較大的圖像中,可能會有身體重疊的問題存在,進而會造成識別到的行人數量漏檢或將同一人多次統計的錯檢情況,進而降低人流量統計的準確性。 With the development of science and technology, face recognition technology is widely used in different fields, for example, in the company as a commuter swipe card, in the communication field as an unlock password for mobile devices, etc. Based on face recognition technology, the flow of people in the image can be counted. In images with a high degree of crowd density, there may be a problem of overlapping bodies, which will cause the number of identified pedestrians to be missed or the same person to be counted multiple times, thus reducing the accuracy of people flow statistics. .
本發明的主要目的是提供一種基於人臉識別的即時行人統計方法以及裝置,旨在解決現有技術中的對密集程度較大的圖像進行人臉識別存在的漏檢或誤檢的問題。 The main purpose of the present invention is to provide a real-time pedestrian counting method and device based on face recognition, which aims to solve the problem of missed detection or false detection in the prior art in the face recognition of denser images.
一種基於人臉識別的即時行人統計方法,所述即時行人統計方法包括:獲取待檢測視頻;提取所述待檢測視頻中具有行人的圖像作為檢測圖像;利用第一檢測模型對所述檢測圖像進行行人識別並得到行人檢測框;根據所述行人檢測框執行人臉識別操作並提取人臉檢測框; 利用第二檢測模型對所述人臉檢測框內的人臉圖像進行去重操作;判斷資料庫中是否存在相同所述人臉圖像;在所述資料庫中不存在相同所述人臉圖像時,對所述人臉圖像進行編號並基於指定演算法對所述人臉圖像的軌跡進行追蹤;判斷所述軌跡是否經過指定區域;在所述軌跡經過指定區域時,將所述人臉圖像作為目標物件並統計所述目標物件的數量;所述即時行人統計方法還包括:利用第三檢測模型檢測所述人臉檢測框內的人臉特徵參數;所述利用第三檢測模型檢測所述人臉檢測框內的人臉特徵參數的步驟包括:提取所述人臉檢測框內的人臉特徵點;根據所述人臉特徵點對所述人臉檢測框內的所述人臉圖像進行角度校正;判斷所述人臉檢測框的面積是否大於預定面積;在所述人臉檢測框的面積大於所述預定面積時,基於第一函數獲取人臉模糊度;判斷所述人臉模糊度是否大於模糊度閾值;在所述人臉模糊度大於所述模糊度閾值時,利用所述第三檢測模型檢測所述人臉檢測框內的人臉圖像的年齡和性別。 A real-time pedestrian statistical method based on face recognition, the real-time pedestrian statistical method includes: acquiring a video to be detected; extracting an image of a pedestrian in the video to be detected as a detection image; using a first detection model to detect the detection Perform pedestrian recognition on the image and obtain a pedestrian detection frame; perform a face recognition operation and extract a face detection frame according to the pedestrian detection frame; Use the second detection model to deduplicate the face image in the face detection frame; determine whether the same face image exists in the database; the same face does not exist in the database During the image, number the face image and track the trajectory of the face image based on the specified algorithm; determine whether the trajectory passes through the specified area; when the trajectory passes through the specified area, using the face image as a target object and counting the number of the target objects; the real-time pedestrian counting method further comprises: using a third detection model to detect the face feature parameters in the face detection frame; The step of detecting the face feature parameters in the face detection frame by the detection model includes: extracting face feature points in the face detection frame; performing angle correction on the face image; judging whether the area of the face detection frame is larger than a predetermined area; when the area of the face detection frame is larger than the predetermined area, obtain the face ambiguity based on the first function; determine Whether the ambiguity of the face is greater than the ambiguity threshold; when the ambiguity of the face is greater than the ambiguity threshold, the third detection model is used to detect the age and the age of the face image in the face detection frame. gender.
優選地,所述第一檢測模型為基於YoloV3演算法實現的深度學習人體檢測模型。 Preferably, the first detection model is a deep learning human detection model implemented based on the YoloV3 algorithm.
優選地,所述人臉識別操作為基於電腦視覺庫實現。 Preferably, the face recognition operation is implemented based on a computer vision library.
優選地,所述第二檢測模型為超解析度測試序列模型。 Preferably, the second detection model is a super-resolution test sequence model.
優選地,所述指定演算法為KCF高速跟蹤演算法。 Preferably, the specified algorithm is a KCF high-speed tracking algorithm.
優選地,所述第一函數為拉普拉斯運算元函數。 Preferably, the first function is a Laplace operator function.
優選地,所述第三檢測模型為超解析度測試序列模型。 Preferably, the third detection model is a super-resolution test sequence model.
此外,為了實現上述目的,本發明還提出一種基於人臉識別的即時行人統計裝置,所述即時行人統計裝置包括處理器和記憶體,所述處理器用於執行所述記憶體中存儲的電腦程式時實現如下步驟:獲取待檢測視頻;提取所述待檢測視頻中具有行人的圖像作為檢測圖像;利用第一檢測模型對所述檢測圖像進行行人識別並得到行人檢測框;根據所述行人檢測框執行人臉識別操作並提取人臉檢測框;利用第二檢測模型對所述人臉檢測框內的人臉圖像進行去重操作;判斷資料庫中是否存在相同的所述人臉圖像;在所述資料庫中不存在相同的所述人臉圖像時,對所述人臉圖像進行編號並基於指定演算法對所述人臉圖像的軌跡進行追蹤;判斷所述軌跡是否經過指定區域;在所述軌跡經過指定區域時,將所述人臉圖像作為目標物件並統計所述目標物件的數量;所述即時行人統計方法還包括:利用第三檢測模型檢測所述人臉檢測框內的人臉特徵參數;所述利用第三檢測模型檢測所述人臉檢測框內的人臉特徵參數的步驟包括:提取所述人臉檢測框內的人臉特徵點;根據所述人臉特徵點對所述人臉檢測框內的所述人臉圖像進行角度校正;判斷所述人臉檢測框的面積是否大於預定面積; 在所述人臉檢測框的面積大於所述預定面積時,基於第一函數獲取人臉模糊度;判斷所述人臉模糊度是否大於模糊度閾值;在所述人臉模糊度大於所述模糊度閾值時,利用所述第三檢測模型檢測所述人臉檢測框內的人臉圖像的年齡和性別。 In addition, in order to achieve the above object, the present invention also provides a real-time pedestrian counting device based on face recognition, the real-time pedestrian counting device includes a processor and a memory, and the processor is used to execute a computer program stored in the memory. The following steps are implemented: acquiring the video to be detected; extracting the image with pedestrians in the video to be detected as the detection image; using the first detection model to perform pedestrian recognition on the detected image and obtain a pedestrian detection frame; The pedestrian detection frame performs the face recognition operation and extracts the face detection frame; uses the second detection model to carry out the deduplication operation on the face image in the face detection frame; judges whether there is the same described face in the database image; when the same face image does not exist in the database, number the face image and track the trajectory of the face image based on a specified algorithm; judge the face image Whether the track passes through the designated area; when the track passes through the designated area, the face image is used as a target object and the number of the target objects is counted; the real-time pedestrian counting method further includes: using a third detection model to detect all the face feature parameters in the face detection frame; the step of detecting the face feature parameters in the face detection frame by using the third detection model includes: extracting the face feature points in the face detection frame; Perform angle correction on the face image in the face detection frame according to the face feature points; determine whether the area of the face detection frame is larger than a predetermined area; When the area of the face detection frame is greater than the predetermined area, obtain a face blur based on the first function; determine whether the face blur is greater than a blur threshold; when the face blur is greater than the blur When the degree threshold is set, the third detection model is used to detect the age and gender of the face image in the face detection frame.
上述基於人臉識別的即時行人統計方法以及裝置,藉由對提取到的人臉檢測框進行去除重複操作,可減少行人重疊或部分重疊時的誤判,進而提高行人統計的準確性。 The above-mentioned real-time pedestrian counting method and device based on face recognition can reduce the misjudgment when pedestrians overlap or partially overlap by performing repetitive operations on the extracted face detection frames, thereby improving the accuracy of pedestrian counting.
100:即時行人統計裝置 100: Instant Pedestrian Statistics Device
102:記憶體 102: Memory
103:處理器 103: Processor
104:通信匯流排 104: Communication busbar
106:圖像採集設備 106: Image acquisition equipment
1:即時行人統計系統 1: Instant Pedestrian Statistics System
2:操作系統 2: Operating system
10:獲取模組 10: Get mods
20:提取模組 20: Extract the module
30:檢測模組 30: Detection module
40:判斷模組 40: Judgment Module
50:追蹤模組 50: Tracking Mods
60:統計模組 60: Statistics Module
S10-S19:步驟 S10-S19: Steps
圖1為本發明即時行人統計裝置的功能模組圖。 FIG. 1 is a functional module diagram of the instant pedestrian counting device of the present invention.
圖2為圖1中所述即時行人統計系統的功能模組圖。 FIG. 2 is a functional module diagram of the instant pedestrian counting system described in FIG. 1 .
圖3為本發明的基於人臉識別的即時行人統計方法的流程圖。 FIG. 3 is a flow chart of the instant pedestrian counting method based on face recognition according to the present invention.
圖4為圖3中步驟S19的細化流程示意圖。 FIG. 4 is a schematic diagram of the refinement flow of step S19 in FIG. 3 .
為了使本技術領域的人員更好地理解本發明方案,下面將結合本發明實施例中的附圖,對本發明實施例中的技術方案進行清楚、完整地描述,顯然,所描述的實施例僅僅是本發明一部分的實施例,而不是全部的實施例。基於本發明中的實施例,本領域普通技術人員在沒有做出創造性勞動前提下所獲得的所有其他實施例,都應當屬於本發明保護的範圍。 In order to make those skilled in the art better understand the solutions of the present invention, the technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only Embodiments are part of the present invention, but not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by persons of ordinary skill in the art without creative efforts shall fall within the protection scope of the present invention.
本發明的說明書及上述附圖中的術語「第一」、「第二」和「第三」等是用於區別不同物件,而非用於描述特定順序。此外,術語「包括」以及它們任何變形,意圖在於覆蓋不排他的包含。例如包含了一系列步驟 或模組的過程、方法、系統、產品或設備沒有限定於已列出的步驟或模組,而是可選地還包括沒有列出的步驟或模組,或可選地還包括對於這些過程、方法、產品或設備固有的其它步驟或模組。 The terms "first", "second" and "third" in the description of the present invention and the above-mentioned drawings are used to distinguish different items, rather than to describe a specific order. Furthermore, the term "comprising" and any variations thereof are intended to cover non-exclusive inclusion. For example, it contains a series of steps or modules of processes, methods, systems, products or devices are not limited to the listed steps or modules, but optionally also include unlisted steps or modules, or optionally also include , method, product or other steps or modules inherent in the device.
下面結合附圖對本發明基於即時行人統計裝置以及即時行人統計方法的具體實施方式進行說明。 The specific embodiments of the present invention based on the instant pedestrian counting device and the instant pedestrian counting method will be described below with reference to the accompanying drawings.
請參照圖1,本發明提供一種即時行人統計裝置100。所述即時行人統計裝置100包括記憶體102、處理器103、通信匯流排104以及至少一個圖像採集設備106。所述即時行人統計裝置100可與伺服器之間可藉由預設協定進行通訊。優選地,所述預設協議包括,但不限於以下任意一種:HTTP協定(Hyper Text Transfer Protocol,超文字傳輸協定)、HTTPS協議(Hyper Text Transfer Protocol over Secure Socket Layer,以安全為目標的HTTP協定)等。所述伺服器可以是單一的伺服器,也可以為由幾個功能伺服器共同組成的伺服器群。在本發明的至少一個實施例中,所述即時行人統計裝置100可以為是任意具有網路連接功能的終端,例如,所述終端設備可以為個人電腦、平板電腦、智慧手機、個人數位助理(Personal Digital Assistant,PDA)、遊戲機、互動式網路電視(Internet Protocol Television,IPTV)、智慧式穿戴式設備、導航裝置等等的可移動設備,或者臺式電腦、數位TV等等固定設備。
Referring to FIG. 1 , the present invention provides a real-time
所述即時行人統計裝置100根據獲取待檢測視頻並從中提取具有行人的圖像作為檢測圖像。所述即時行人統計裝置100進一步地利用第一檢測模型對所述檢測圖像進行行人識別並得到行人檢測框並對所述行人檢測框進行人臉識別獲得人臉檢測框。所述即時行人統計裝置100進一步地利用第二檢測模型對所述人臉檢測框內的人臉圖像進行去重操作並在所述人臉圖像對應的軌跡未被記錄時基於指定演算法對所述人臉圖像的軌跡進行追
蹤。所述即時行人統計裝置100進一步地在所述軌跡經過指定區域時將所述人臉圖像作為目標物件並統計所述目標物件的數量並輸出。
The instant
所述記憶體102用於存儲程式碼。所述記憶體102可以是積體電路中沒有實物形式的具有存儲功能的電路,如記憶體條、TF卡(Trans-flash Card)、智慧媒體卡(smart media card)、安全數位卡(secure digital card)、快閃記憶體卡(flash card)等儲存設備。所述記憶體102可藉由所述通信匯流排104與所述處理器103進行資料通信。所述記憶體102中可以包括操作系統2以及即時行人統計系統1。所述記憶體102內還可以存儲有資料庫。所述資料庫中用於存儲多個編碼互不相同的人臉圖像。其中所述編碼可以為第一檢測模型輸出的行人特徵向量。
The
所述操作系統2是管理和控制即時行人統計裝置硬體和軟體資源的程式,支援即時行人統計系統以及其它軟體和/或程式的運行。
The
所述處理器103可以包括一個或者多個微處理器、數位信號處理器(DSP,Digtial Signal Processor)。所述處理器103可調用所述記憶體102中存儲的程式碼以執行相關的功能。例如,圖2中所述的模組10-60是存儲在所述記憶體102中的程式碼,並由所述處理器103所執行,以實現一種基於所述高精地圖3的即時行人統計方法。所述處理器103又稱中央處理器(CPU,Central Processing Unit),是一塊超大規模的積體電路,是運算核心(Core)和控制核心(Control Unit)。
The
所述通信匯流排104與所述記憶體102與所述處理器103進行資料通信。
The
所述圖像採集設備106用於拍攝圖像。在本發明的至少一個實施例中,所述圖像採集設備106可設置於所述即時行人統計裝置100內,還可以與所述即時行人統計裝置100分離設置,例如設置於路燈下方的監控攝像
頭,看板上的監控攝像頭,或者廣場標誌物上的監控攝像頭等,但並不局限於此。所述圖像採集設備106還可在所述處理器103的控制下轉動,以實現調整拍攝角度。
The
請一併參閱圖2,其為所述即時行人統計系統1的功能模組示意圖。所述即時行人統計系統1包括:獲取模組10,用於獲取待檢測視頻。
Please also refer to FIG. 2 , which is a schematic diagram of functional modules of the real-time
在本發明的至少一個實施方式中,所述待檢測視頻可由所述圖像採集設備106採集得到,也可以從伺服器內獲取得到。所述待檢測視頻由多幀圖像構成。
In at least one embodiment of the present invention, the video to be detected can be acquired by the
提取模組20,用於提取所述待檢測視頻中具有行人的圖像作為檢測圖像。
The
檢測模組30,用於利用第一檢測模型對所述檢測圖像進行行人識別並得到行人檢測框。
The
在本發明的至少一個實施方式中,所述第一檢測模型為深度學習人體檢測模型,其基於YoloV3演算法實現。 In at least one embodiment of the present invention, the first detection model is a deep learning human detection model, which is implemented based on the YoloV3 algorithm.
所述提取模組20進一步地根據所述行人檢測框執行人臉識別操作並提取人臉檢測框。
The
在本發明的至少一個實施方式中,所述人臉識別操作為基於電腦視覺庫(OpenCV)實現。 In at least one embodiment of the present invention, the face recognition operation is implemented based on a computer vision library (OpenCV).
所述檢測模組30進一步地利用第二檢測模型對所述人臉檢測框內的人臉圖像進行去重操作。
The
在本發明的至少一個實施方式中,所述第二檢測模型為卷積神經網路模型,例如,超解析度測試序列(Visual GeometryGroup,VGG)模型。 其中,所述第二檢測模型VGG具有16層權重層,其中包括13個卷積層,3個全連接層以及5個池化層。所述權重層為具有權重係數。 In at least one embodiment of the present invention, the second detection model is a convolutional neural network model, for example, a super-resolution test sequence (Visual Geometry Group, VGG) model. The second detection model VGG has 16 weight layers, including 13 convolution layers, 3 fully connected layers and 5 pooling layers. The weight layer has weight coefficients.
判斷模組40,用於判斷資料庫中是否存在相同的所述人臉圖像。
The judging
在本發明的至少一個實施方式中,可藉由所述第一檢測模型輸出行人特徵向量,以所述行人特徵向量作為編號在所述資料庫內進行查找。 In at least one embodiment of the present invention, a pedestrian feature vector may be output by the first detection model, and the pedestrian feature vector may be used as a serial number for searching in the database.
追蹤模組50,用於在所述資料庫中不存在相同的所述人臉圖像時對所述人臉圖像進行編號並基於指定演算法對所述人臉圖像的軌跡進行追蹤。
The
在本發明的至少一個實施方式中,所述指定演算法為KCF(Kernelized Correlation Filters)高速跟蹤演算法。 In at least one embodiment of the present invention, the specified algorithm is a KCF (Kernelized Correlation Filters) high-speed tracking algorithm.
所述判斷模組40進一步地所述軌跡是否經過指定區域。
The judging
統計模組60,用於在在所述軌跡經過所述指定區域時將所述人臉圖像作為目標物件並統計所述目標物件的數量。
The
所述檢測模組30進一步地利用第三檢測模型檢測所述人臉檢測框內的人臉特徵參數。
The
所述檢測模組30進一步地提取所述人臉檢測框內的人臉特徵點,根據所述人臉特徵點對所述人臉檢測框內的所述人臉圖像進行角度校正,並判斷所述人臉檢測框的面積是否大於預定面積。在所述人臉檢測框的面積大於所述預定面積時,所述檢測模組30進一步地基於第一函數獲取人臉模糊度並判斷所述人臉模糊度是否大於模糊度閾值。在所述人臉模糊度大於所述模糊度閾值時,所述檢測模組30進一步地利用所述第三檢測模型檢測所述人臉檢測框內的人臉圖像的年齡和性別。
The
在本發明的至少一個實施方式中,所述人臉特徵點為利用Dlib庫進行五點Landmarks。所述第一函數為拉普拉斯(Laplacian)運算元函數。所述第三檢測模型為VGG網路模型。 In at least one embodiment of the present invention, the facial feature points are five-point Landmarks using the Dlib library. The first function is a Laplacian operator function. The third detection model is a VGG network model.
上述所述即時行人統計裝置100,藉由對提取到的人臉檢測框進行去除重複操作,可減少行人重疊或部分重疊時的誤判,進而提高行人統計的準確性。同時,藉由軌跡追蹤以及臉部分析實現人臉性格和年齡的統計,可實現對行人的年齡和性別的精確分析。
The above-mentioned real-time
請參閱圖3,其為即時行人統計方法的流程圖。所述即時行人統計方法應用於具有所述自動閃避系統1的所述即時行人統計裝置100中。所述即時行人統計裝置100還可以包括圖1或圖2更多或更少的其他硬體或者軟體,或者不同的部件設置方式。所述即時行人統計裝置100可提供一視覺化介面。所述視覺化介面用於向使用者提供人機交互介面,使用者可以在藉由手機或電腦等電子設備連接到所述即時行人統計裝置100。
Please refer to FIG. 3 , which is a flow chart of the instant pedestrian counting method. The instant pedestrian counting method is applied to the instant
所述即時行人統計裝置100基於所述處理器103執行存儲在所述記憶體102上的獲取模組10、提取模組20、檢測模組30、判斷模組40、追蹤模組50以及統計模組60,並且與圖像採集設備106可通信地接合來執行所述即時行人統計方法。
The real-time
S10、所述獲取模組10獲取待檢測視頻。
S10. The obtaining
在本發明的至少一個實施方式中,所述待檢測視頻可由所述圖像採集設備106採集得到,也可以從伺服器內獲取得到。所述待檢測視頻由多幀圖像構成。
In at least one embodiment of the present invention, the video to be detected can be acquired by the
S11、所述提取模組20提取所述待檢測視頻中具有行人的圖像作為檢測圖像。
S11. The
S12、所述檢測模組30利用第一檢測模型對所述檢測圖像進行行人識別並得到行人檢測框。
S12. The
在本發明的至少一個實施方式中,所述第一檢測模型為深度學習人體檢測模型,其基於YoloV3演算法實現。 In at least one embodiment of the present invention, the first detection model is a deep learning human detection model, which is implemented based on the YoloV3 algorithm.
S13、所述提取模組20根據所述行人檢測框執行人臉識別操作並提取人臉檢測框。
S13. The
在本發明的至少一個實施方式中,所述人臉識別操作為基於電腦視覺庫(OpenCV)實現。 In at least one embodiment of the present invention, the face recognition operation is implemented based on a computer vision library (OpenCV).
S14、所述檢測模組30利用第二檢測模型對所述人臉檢測框內的人臉圖像進行去重操作。
S14. The
在本發明的至少一個實施方式中,所述第二檢測模型為卷積神經網路模型,例如,超解析度測試序列(Visual GeometryGroup,VGG)模型。其中,所述第二檢測模型具有16層權重層,其中包括13個卷積層,3個全連接層以及5個池化層。所述權重層為具有權重係數。 In at least one embodiment of the present invention, the second detection model is a convolutional neural network model, for example, a super-resolution test sequence (Visual Geometry Group, VGG) model. The second detection model has 16 weight layers, including 13 convolutional layers, 3 fully connected layers and 5 pooling layers. The weight layer has weight coefficients.
S15、所述判斷模組40判斷資料庫中是否存在相同的所述人臉圖像。
S15. The judging
在本發明的至少一個實施方式中,可藉由所述第一檢測模型輸出行人特徵向量,以所述行人特徵向量作為編號在所述資料庫內進行查找。 In at least one embodiment of the present invention, a pedestrian feature vector may be output by the first detection model, and the pedestrian feature vector may be used as a serial number for searching in the database.
在所述資料庫中存在相同的所述人臉圖像時,流程結束。 When the same face image exists in the database, the process ends.
S16、所述追蹤模組50在所述資料庫中不存在相同的所述人臉圖像時對所述人臉圖像進行編號並基於指定演算法對所述人臉圖像的軌跡進行追蹤。
S16, the
S17、所述判斷模組40判斷所述軌跡是否經過指定區域。
S17. The judging
在所述軌跡未經過所述指定區域時,流程結束。 When the trajectory does not pass through the designated area, the process ends.
S18、所述統計模組60在所述軌跡經過所述指定區域時將所述人臉圖像作為目標物件並統計所述目標物件的數量。
S18. The
S19、所述檢測模組30利用第三檢測模型檢測所述人臉圖像的特徵參數。
S19. The
請一併參閱圖4,在本發明的至少一個實施方式中,所述檢測模組30利用第三檢測模型檢測所述人臉圖像的特徵參數的步驟進一步包括:S191、所述檢測模組30提取所述人臉檢測框內的人臉特徵點;S192、所述檢測模組30根據所述人臉特徵點對所述人臉檢測框內的所述人臉圖像進行角度校正;S193、所述檢測模組30判斷所述人臉檢測框的面積是否大於預定面積;S194、所述檢測模組30在所述人臉檢測框的面積大於所述預定面積時基於第一函數獲取人臉模糊度;S195、所述檢測模組30判斷所述人臉模糊度是否大於模糊度閾值;S196、所述檢測模組30在所述人臉模糊度大於所述模糊度閾值時利用所述第三檢測模型檢測所述人臉圖像的年齡和性別。
Please also refer to FIG. 4 , in at least one embodiment of the present invention, the step of the
在所述人臉檢測框的面積小於等於所述預定面積或所述人臉模糊度小於等於所述模糊度閾值時,返回步驟S11。 When the area of the face detection frame is less than or equal to the predetermined area or the blurriness of the face is less than or equal to the blurriness threshold, return to step S11.
在本發明的至少一個實施方式中,所述人臉特徵點為利用Dlib庫進行五點Landmarks。所述第一函數為拉普拉斯(Laplacian)運算元函數。所述第三檢測模型為VGG網路模型。 In at least one embodiment of the present invention, the facial feature points are five-point Landmarks using the Dlib library. The first function is a Laplacian operator function. The third detection model is a VGG network model.
上述即時行人統計方法,藉由對提取到的人臉檢測框進行去除重複操作,可減少行人重疊或部分重疊時的誤判,進而提高行人統計的準確 性。同時,藉由軌跡追蹤以及臉部分析實現人臉性格和年齡的統計,可實現對行人的年齡和性別的精確分析。 The above real-time pedestrian counting method can reduce the misjudgment when pedestrians overlap or partially overlap by deduplicating the extracted face detection frame, thereby improving the accuracy of pedestrian counting. sex. At the same time, the statistics of face personality and age can be realized through trajectory tracking and face analysis, and accurate analysis of the age and gender of pedestrians can be achieved.
需要說明的是,對於前述的各方法實施例,為了簡單描述,故將其都表述為一系列的動作組合,但是本領域技術人員應該知悉,本發明並不受所描述的動作順序的限制,因為依據本發明,某些步驟可以採用其他順序或者同時進行。其次,本領域技術人員也應該知悉,說明書中所描述的實施例均屬於優選實施例,所涉及的動作和模組並不一定是本發明所必須的。 It should be noted that, for the sake of simple description, the foregoing method embodiments are all expressed as a series of action combinations, but those skilled in the art should know that the present invention is not limited by the described action sequence. As in accordance with the present invention, certain steps may be performed in other orders or simultaneously. Secondly, those skilled in the art should also know that the embodiments described in the specification are all preferred embodiments, and the actions and modules involved are not necessarily required by the present invention.
在本申請所提供的幾個實施例中,應該理解到,所揭露的裝置,可藉由其它的方式實現。例如,以上所描述的裝置實施例僅僅是示意性的,例如所述模組的劃分,僅僅為一種邏輯功能劃分,實際實現時可以有另外的劃分方式,例如多個模組或元件可以結合或者可以集成到另一個系統,或一些特徵可以忽略,或不執行。另一點,所顯示或討論的相互之間的耦合或直接耦合或通信連接可以是藉由一些介面,裝置或模組的間接耦合或通信連接,可以是電性或其它的形式。 In the several embodiments provided in this application, it should be understood that the disclosed apparatus may be implemented in other manners. For example, the device embodiments described above are only illustrative. For example, the division of the modules is only a logical function division. In actual implementation, there may be other division methods. For example, multiple modules or components may be combined or Can be integrated into another system, or some features can be ignored, or not implemented. On the other hand, the shown or discussed mutual coupling or direct coupling or communication connection may be through some interfaces, indirect coupling or communication connection of devices or modules, and may be electrical or other forms.
所述作為分離部件說明的模組可以是或者也可以不是物理上分開的,作為模組顯示的部件可以是或者也可以不是物理模組,即可以位於一個地方,或者也可以分佈到多個網路模組上。可以根據實際的需要選擇其中的部分或者全部模組來實現本實施例方案的目的。 The modules described as separate components may or may not be physically separated, and the components displayed as modules may or may not be physical modules, that is, they may be located in one place, or may be distributed to multiple networks. on the road module. Some or all of the modules may be selected according to actual needs to achieve the purpose of the solution in this embodiment.
另外,在本發明的各個實施例中的各功能模組可以集成在一個處理器中,也可以是各個模組單獨物理存在,也可以兩個或兩個以上模組集成在一個模組中。上述集成的模組既可以採用硬體的形式實現,也可以採用軟體功能模組的形式實現。 In addition, each functional module in each embodiment of the present invention may be integrated into one processor, or each module may exist physically alone, or two or more modules may be integrated into one module. The above-mentioned integrated modules can be implemented in the form of hardware, or can be implemented in the form of software function modules.
所述集成的模組如果以軟體功能模組的形式實現並作為獨立的產品銷售或使用時,可以存儲在一個電腦可讀取存儲介質中。基於這樣的理解,本發明的技術方案本質上或者說對現有技術做出貢獻的部分或者該技術方案的全部或部分可以以軟體產品的形式體現出來,該電腦軟體產品存儲在一個存儲介質中,包括若干指令用以使得一台電腦設備(可為個人電腦、伺服器或者網路設備等)執行本發明各個實施例所述方法的全部或部分步驟。 If the integrated module is implemented in the form of a software function module and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention is essentially or the part that contributes to the prior art or all or part of the technical solution can be embodied in the form of a software product, and the computer software product is stored in a storage medium, Several instructions are included to cause a computer device (which may be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of the present invention.
還需要說明的是,在本文中,術語「包括」、「包含」或者其任何其他變體意在涵蓋非排他性的包含,從而使得包括一系列要素的過程、方法、物品或者裝置不僅包括那些要素,而且還包括沒有明確列出的其他要素,或者是還包括為這種過程、方法、物品或者裝置所固有的要素。在沒有更多限制的情況下,由語句「包括一個......」限定的要素,並不排除在包括該要素的過程、方法、物品或者裝置中還存在另外的相同要素。 It should also be noted that, herein, the terms "comprising", "comprising" or any other variation thereof are intended to encompass non-exclusive inclusion, such that a process, method, article or device comprising a series of elements includes not only those elements , but also other elements not expressly listed or inherent to such a process, method, article or apparatus. Without further limitation, an element qualified by the phrase "comprising a..." does not preclude the presence of additional identical elements in a process, method, article or apparatus that includes the element.
以上所述,以上實施例僅用以說明本發明的技術方案,而非對其限制;儘管參照前述實施例對本發明進行了詳細的說明,本領域的普通技術人員應當理解:其依然可以對前述各實施例所記載的技術方案進行修改,或者對其中部分技術特徵進行等同替換;而這些修改或者替換,並不使相應技術方案的本質脫離本發明各實施例技術方案的範圍。 As mentioned above, the above embodiments are only used to illustrate the technical solutions of the present invention, but not to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand: The technical solutions described in the embodiments are modified, or some technical features thereof are equivalently replaced; and these modifications or replacements do not make the essence of the corresponding technical solutions depart from the scope of the technical solutions of the embodiments of the present invention.
綜上所述,本發明符合發明專利要件,爰依法提出專利申請。惟,以上所述者僅為本發明之較佳實施方式,舉凡熟悉本案技藝之人士,在爰依本案創作精神所作之等效修飾或變化,皆應包含於以下之申請專利範圍內。 To sum up, the present invention complies with the requirements of an invention patent, and a patent application can be filed in accordance with the law. However, the above descriptions are only the preferred embodiments of the present invention, and for those who are familiar with the techniques of this case, equivalent modifications or changes made in accordance with the creative spirit of this case should all be included in the scope of the following patent application.
S10-S19:步驟 S10-S19: Steps
Claims (8)
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
TW109143733A TWI764425B (en) | 2020-12-10 | 2020-12-10 | Real time pedestrian statistical method based on face identification, and apparatus thereof |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
TW109143733A TWI764425B (en) | 2020-12-10 | 2020-12-10 | Real time pedestrian statistical method based on face identification, and apparatus thereof |
Publications (2)
Publication Number | Publication Date |
---|---|
TWI764425B true TWI764425B (en) | 2022-05-11 |
TW202223731A TW202223731A (en) | 2022-06-16 |
Family
ID=82594065
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
TW109143733A TWI764425B (en) | 2020-12-10 | 2020-12-10 | Real time pedestrian statistical method based on face identification, and apparatus thereof |
Country Status (1)
Country | Link |
---|---|
TW (1) | TWI764425B (en) |
Citations (7)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
TW201629851A (en) * | 2015-01-12 | 2016-08-16 | 芋頭科技(杭州)有限公司 | A system and a method for image recognition |
CN106156688A (en) * | 2015-03-10 | 2016-11-23 | 上海骏聿数码科技有限公司 | A kind of dynamic human face recognition methods and system |
TWM558943U (en) * | 2017-11-22 | 2018-04-21 | Aiwin Technology Co Ltd | Intelligent image information and big data analysis system using deep-learning technology |
WO2019033574A1 (en) * | 2017-08-17 | 2019-02-21 | 平安科技(深圳)有限公司 | Electronic device, dynamic video face recognition method and system, and storage medium |
CN109598211A (en) * | 2018-11-16 | 2019-04-09 | 恒安嘉新(北京)科技股份公司 | A kind of real-time dynamic human face recognition methods and system |
TW202030637A (en) * | 2019-01-31 | 2020-08-16 | 大陸商北京市商湯科技開發有限公司 | Method and device, electronic equipment for face image recognition and storage medium thereof |
CN111738215A (en) * | 2020-07-21 | 2020-10-02 | 深圳市视美泰技术股份有限公司 | Human body temperature measuring method and computer equipment |
-
2020
- 2020-12-10 TW TW109143733A patent/TWI764425B/en active
Patent Citations (7)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
TW201629851A (en) * | 2015-01-12 | 2016-08-16 | 芋頭科技(杭州)有限公司 | A system and a method for image recognition |
CN106156688A (en) * | 2015-03-10 | 2016-11-23 | 上海骏聿数码科技有限公司 | A kind of dynamic human face recognition methods and system |
WO2019033574A1 (en) * | 2017-08-17 | 2019-02-21 | 平安科技(深圳)有限公司 | Electronic device, dynamic video face recognition method and system, and storage medium |
TWM558943U (en) * | 2017-11-22 | 2018-04-21 | Aiwin Technology Co Ltd | Intelligent image information and big data analysis system using deep-learning technology |
CN109598211A (en) * | 2018-11-16 | 2019-04-09 | 恒安嘉新(北京)科技股份公司 | A kind of real-time dynamic human face recognition methods and system |
TW202030637A (en) * | 2019-01-31 | 2020-08-16 | 大陸商北京市商湯科技開發有限公司 | Method and device, electronic equipment for face image recognition and storage medium thereof |
CN111738215A (en) * | 2020-07-21 | 2020-10-02 | 深圳市视美泰技术股份有限公司 | Human body temperature measuring method and computer equipment |
Also Published As
Publication number | Publication date |
---|---|
TW202223731A (en) | 2022-06-16 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
US11074436B1 (en) | Method and apparatus for face recognition | |
WO2019120115A1 (en) | Facial recognition method, apparatus, and computer apparatus | |
CN106203305B (en) | Face living body detection method and device | |
CN111429483A (en) | High-speed cross-camera multi-target tracking method, system, device and storage medium | |
WO2019071664A1 (en) | Human face recognition method and apparatus combined with depth information, and storage medium | |
CN111738120B (en) | Character recognition method, character recognition device, electronic equipment and storage medium | |
CN114677607A (en) | Real-time pedestrian counting method and device based on face recognition | |
CN108108711B (en) | Face control method, electronic device and storage medium | |
TWI704505B (en) | Face recognition system, establishing data method for face recognition, and face recognizing method thereof | |
WO2023173646A1 (en) | Expression recognition method and apparatus | |
CN109684993B (en) | Face recognition method, system and equipment based on nostril information | |
CN109902550A (en) | The recognition methods of pedestrian's attribute and device | |
CN112529939A (en) | Target track matching method and device, machine readable medium and equipment | |
CN111476132A (en) | Video scene recognition method and device, electronic equipment and storage medium | |
CN114783061A (en) | Smoking behavior detection method, device, equipment and medium | |
CN108255352B (en) | Multi-touch implementation method and system in projection interaction system | |
TWI764425B (en) | Real time pedestrian statistical method based on face identification, and apparatus thereof | |
Sandhiyasa et al. | Real Time Face Recognition for Mobile Application Based on Mobilenetv2 | |
CN115457623A (en) | Stay time duration detection method and device and server | |
WO2003056501A1 (en) | Methods and apparatus for face recognition | |
WO2022140879A1 (en) | Identity recognition method, terminal, server, and system | |
CN113902030A (en) | Behavior identification method and apparatus, terminal device and storage medium | |
CN113947781A (en) | Lost child identification method, lost child identification system, terminal equipment and storage medium | |
CN111461104B (en) | Visual recognition method, device, equipment and storage medium | |
CN108694347B (en) | Image processing method and device |