TWI611374B - Gender and age identification method for vertical image flow counting - Google Patents
Gender and age identification method for vertical image flow counting Download PDFInfo
- Publication number
- TWI611374B TWI611374B TW106114737A TW106114737A TWI611374B TW I611374 B TWI611374 B TW I611374B TW 106114737 A TW106114737 A TW 106114737A TW 106114737 A TW106114737 A TW 106114737A TW I611374 B TWI611374 B TW I611374B
- Authority
- TW
- Taiwan
- Prior art keywords
- image
- age
- gender
- shoulder
- information
- Prior art date
Links
Landscapes
- Image Analysis (AREA)
- Measurement Of The Respiration, Hearing Ability, Form, And Blood Characteristics Of Living Organisms (AREA)
Description
本發明屬於一種垂直式影像人流計數之性別與年齡辨識方法,尤指一種可保護來客分析、人流計數與3D影像辨識技術。 The invention belongs to a method for identifying gender and age of vertical image flow counting, in particular to a technology for protecting visitor analysis, flow counting and 3D image recognition.
目前其他公司的人流計數產品遭遇僅能計算人數不能分析其客層屬性之缺失,若需達成此功能,是需要透過多個硬體設備輔助達成,硬體成本提升導致推廣不易。 At present, other companies' flow counting products encounter only the number of people can not analyze the lack of their customer layer attributes. If this function is required, it needs to be assisted by multiple hardware devices, and the hardware cost increase will not be easy to promote.
本案發明人鑑於上述習用方式所衍生的各項缺點,乃亟思加以改良創新,並經多年苦心孤詣潛心研究後,終於成功研發完成本垂直式影像人流計數之性別與年齡辨識方法。 In view of the shortcomings derived from the above-mentioned conventional methods, the inventor of the present invention has improved and innovated, and after years of painstaking research, he finally successfully developed the gender and age identification method for the vertical image flow counting.
為達上述目的,本發明提出提供一種垂直式影像人流計數之性別與年齡辨識方法,主要對於高準確度人流計數與客層屬性分析整合產品有著高度需求,並提供一種應用於垂直式人流計數系統之性別年齡辨識方法,結合影像辨識與身高、肩寬、頭身比例之創新性別年齡辨識方法,改善當前垂直式人流計數應用,僅能計算人數不能分析其客層屬性之缺失,透過提供多元客層屬性資訊,提升影像式人流計數系統 之實用性與產品競爭力。 In order to achieve the above object, the present invention provides a gender and age identification method for vertical image flow counting, which is highly demanded for high-accuracy flow counting and customer layer attribute analysis integration products, and provides a vertical flow counting system. Gender age identification method, combined with image recognition and height, shoulder width, head-to-head ratio, innovative gender age identification method, improve the current vertical flow count application, can only calculate the number of people can not analyze the lack of customer layer attributes, by providing multiple customer layer attribute information , enhance the image flow counting system Practicality and product competitiveness.
根據日本人人體尺寸數據庫(人体寸法 )1997-98研究中,其探討男性與女性在身高和肩寬的差異性,共收集男性110人與女性107人,其年齡分布在20~79歲的測試人員,如下表1所示:
我們可以觀察男性與女性在身高與肩寬的資料分布,男性的平均身高為170公分、平均肩寬為40公分;女性的平均身高為158公分、平均肩寬為36公分,也發現男性和女性在身高與肩寬的特徵可以明顯分成兩個群組,為性別辨識識別力高的特徵。 We can observe the distribution of height and shoulder width between men and women. The average height of men is 170 cm and the average shoulder width is 40 cm. The average height of women is 158 cm and the average shoulder width is 36 cm. Men and women are also found. The characteristics of height and shoulder width can be clearly divided into two groups, which are characterized by high recognition and recognition of gender.
如下表2所示:
美國在2010美國海軍陸戰隊人員人體測量調查:方法和總結統計(Anthropometric Survey of U.S.Marine Corps Personnel:Methods and Summary Statistics)的報告,也可以觀察到性別分析與身高、肩寬的關係性。 The US can also observe the relationship between gender analysis and height and shoulder width in the 2010 Anthropometric Survey of U.S. Marine Corps Personnel: Methods and Summary Statistics report.
如下表3所示:
為安德魯.羅密士(Andrew Loomis)在人體結構素描學習書藉中“Figure Drawing for All Its Worth”提出理想人體比例研究,可以觀察到從嬰兒頭身比例為1:3;到幼兒頭身比例的1:5;慢慢到青年、成年的1:7左右,隨著年紀的增長不同年齡層在頭身比例有著明顯的變化,因此若我們可以偵測頭長與身高去計算頭身比例,進而可以分析年齡輸出年齡層辨識結果。 For Andrew. Andrew Loomis's "Figure Drawing for All Its Worth" is a study of the proportion of ideal human body. It can be observed that the ratio from the baby's head is 1:3; to the ratio of the head of the child: 5; slowly to the youth, adult about 1:7, as the age increases, the age group has obvious changes in the head body proportion, so if we can detect the head length and height to calculate the head body ratio, then you can analyze Age output age layer identification results.
一種垂直式影像人流計數之性別與年齡辨識方法,其包括: 步驟一、以3D攝影機,採垂直之俯視方式,擷取人員之影像;步驟二、從人員之影像,得到一3D深度影像資訊及2D彩色影像資訊,並進行影像前處理;步驟三、依3D深度影像資訊,得到頭頂位置、身高、肩膀位置、肩膀區域及肩寬資料;步驟四、以2D彩色影像資訊,進行影像辨識,得到一影像辨識結果;步驟五、依身高、肩寬及影像辨識結果,計算得到人員性別資訊;步驟六、依頭頂、肩膀位置及身高,得到一頭身比例資訊,結合影像辨識結果辨識人員之年齡屬性。 A method for identifying gender and age of vertical image flow counts, comprising: Step 1: Take a 3D camera and take a vertical view to capture the image of the person; Step 2: Obtain a 3D depth image information and 2D color image information from the image of the person, and perform image pre-processing; Step 3: According to 3D Depth image information, get the top position, height, shoulder position, shoulder area and shoulder width data; Step 4, 2D color image information, image recognition, get an image recognition result; Step 5, by height, shoulder width and image recognition As a result, the gender information of the person is calculated; in step 6, according to the head, shoulder position and height, a body proportion information is obtained, and the age attribute of the person is identified by combining the image recognition result.
其中步驟一,是以透過紅外線式三原色深度(RGB-D)感測器安裝於天花板,並以從上而下俯視拍攝取得3D攝影機影像串流;步驟二,是為影像前處理,將3D深度影像資訊與2D彩色影像資訊兩者進行位置對齊校正,再將3D深度影像資訊轉換為灰階影像供後續進行前景物件偵測使用,而2D彩色影像資訊則進行Γ函數(gamma)校正,亦供後續使用;步驟三之3D深度影像資訊,是包含頭頂偵測模組,是以3D深度影像透過自適應式高斯混合模型進行背景建模並分離出前景物件,以橢圓頭形偵測找出前景物件頭頂位置,透過與感測器的距離計算出人員身高資訊;以及肩膀偵測模組,是利用頭部位置以輻射狀向外延伸向下搜尋深度值相同之水平平面,經由橢圓形偵測得到長軸之肩膀兩側位置與肩膀區域,利用肩膀兩側位置計算得到肩寬資訊;步驟四之影像辨識,是以一影像辨識模組利用肩膀區域之2D彩色影像以3D深度影像資 訊濾除背景影像後,先利用肩膀兩側位置做影像正規化,而後計算其灰階特徵或局部二值模式(Local binary patterns,LBP)特徵,並以支持向量機(Support Vector Machines,SVM)、神經網路(Neural Network)或自適應增強(AdaBoost)之機器學習方法做客層屬性分類,產生各類別機率分數值;步驟五之性別資訊,是以性別辨識模組透過身高與肩寬數值以SVM、Neural Network或AdaBoost之機器學習方法得到性別各類別機率分數後,結合影像辨識機率分數,輸出性別辨識結果;步驟六,是包含頭身比例計算模組,是以頭部到肩膀的距離扣除脖子比例常數得到頭長,再以身高減去頭長得到身長,進而計算出頭身比例資訊;以及年齡辨識模組,是利用頭身比例數值以SVM、Neural Network或AdaBoost之機器學習方法得到年齡各類別機率分數後,結合影像辨識機率分數,輸出年齡辨識結果即得以得到年齡屬性,其年齡屬性,是分為三類年齡層,包括幼兒,0至12歲、青年,12至18歲,以及成年18歲以上。 Step 1 is to install the 3D camera image stream through the infrared three-color depth (RGB-D) sensor and to capture the 3D camera image from top to bottom; step 2 is for image pre-processing, 3D depth The image information and the 2D color image information are both aligned and corrected, and the 3D depth image information is converted into grayscale image for subsequent foreground object detection, and the 2D color image information is corrected by gamma function. Subsequent use; Step 3 of the 3D depth image information, including the overhead detection module, is a 3D depth image through the adaptive Gaussian mixture model for background modeling and separation of foreground objects, elliptical head detection to find the foreground The top position of the object calculates the height information of the person through the distance from the sensor; and the shoulder detection module uses the position of the head to radially extend downward to search for a horizontal plane of the same depth value through the elliptical detection. Obtain the position of the shoulders on both sides of the long axis and the shoulder area. Calculate the shoulder width information by using the position on both sides of the shoulder. The image recognition in step 4 is based on an image. Module using the shoulder area 2D color images in 3D image depth resources After filtering the background image, first use the position on both sides of the shoulder to normalize the image, and then calculate its gray-scale features or local binary patterns (LBP) features, and use Support Vector Machines (SVM). The neural network (Neural Network) or the adaptive enhancement (AdaBoost) machine learning method is used to classify the client attributes and generate the probability scores of each category. The gender information in step 5 is based on the height and shoulder width values of the gender identification module. The machine learning method of SVM, Neural Network or AdaBoost obtains the probability scores of each gender category, and combines the image recognition probability score to output the gender identification result; in step 6, it includes the head body proportion calculation module, which is deducted from the head to shoulder distance. The neck ratio constant is obtained from the head length, and then the height is subtracted from the head length to obtain the length of the head, and then the head body ratio information is calculated; and the age identification module uses the machine body learning method of SVM, Neural Network or AdaBoost to obtain the ages. After the class probability score, combined with the image identification probability score, the age identification result is output to obtain the age. Attributes, whose age attributes are divided into three categories, include young children, 0 to 12 years old, youth, 12 to 18 years old, and adults over 18 years of age.
本發明所提供一種垂直式影像人流計數之性別與年齡辨識方法,與其他習用技術相互比較時,更具備下列優點: The invention provides a gender and age identification method for vertical image flow counting, and has the following advantages when compared with other conventional technologies:
1.僅需一台3D深度影像擷取設備,即可提供高準確率之人流計數功能與多元化之客層屬性分析資訊。 1. Only one 3D depth image capture device is needed to provide high-accuracy flow counting function and diversified customer layer attribute analysis information.
2.提出創新之結合人體統計資料如身高、肩寬、頭身比例與影像辨識技術之客層屬性辨識方法。 2. To propose an innovative combination of human statistic data such as height, shoulder width, head-to-body ratio and image recognition technology.
3.滿足業主對於高準確度人流計數與客層屬性分析整合產品之需求,提升影像式人流計數系統之實用性與產品競爭力。 3. To meet the needs of the owner for the integration of high-accuracy flow count and customer-level attribute analysis products, and to improve the practicality and product competitiveness of the image-based flow counting system.
S110~S160‧‧‧流程 S110~S160‧‧‧Process
請參閱有關本發明之詳細說明及其附圖,將可進一步瞭解本發明之技術內容及其目的功效;有關附圖為:圖1為本發明垂直式影像人流計數之性別與年齡辨識方法之流程圖;圖2為本發明垂直式影像人流計數之性別與年齡辨識方法之實施例示意圖。 Please refer to the detailed description of the present invention and the accompanying drawings, and the technical contents of the present invention and the functions thereof can be further understood. The related drawings are: FIG. 1 is a flow chart of the gender and age identification method for vertical image flow counting in the present invention. FIG. 2 is a schematic diagram of an embodiment of a method for identifying gender and age of a vertical image human flow count according to the present invention.
為了使本發明的目的、技術方案及優點更加清楚明白,下面結合附圖及實施例,對本發明進行進一步詳細說明。應當理解,此處所描述的具體實施例僅用以解釋本發明,但並不用於限定本發明。 The present invention will be further described in detail below with reference to the accompanying drawings and embodiments. It is understood that the specific embodiments described herein are merely illustrative of the invention and are not intended to limit the invention.
以下,結合附圖對本發明進一步說明:請參閱圖1所示,為本發明垂直式影像人流計數之性別與年齡辨識方法之流程圖,其包括:步驟一、S110以3D攝影機,採垂直之俯視方式,擷取人員之影像;步驟二、S120從人員之影像,得到一3D深度影像資訊及2D彩色影像資訊,並進行影像前處理;步驟三、S130依3D深度影像資訊,得到頭頂位置、身高、肩膀位置、肩膀區域及肩寬資料;步驟四、S140以2D彩色影像資訊,進行影像辨識,得到一影像辨識結果;步驟五、S150依身高、肩寬及影像辨識結果,計算得到人員性別資訊;步驟六、S160依頭頂、肩膀位置及身高,得到一頭身比例資訊,結合影像辨識結果辨識人員之年齡屬性。 Hereinafter, the present invention will be further described with reference to the accompanying drawings: Referring to FIG. 1 , it is a flowchart of a method for identifying gender and age of a vertical image human flow count according to the present invention, which includes: Step 1 and S110 adopt a 3D camera, and adopt a vertical view. The method is to capture the image of the person; in step 2, the S120 obtains a 3D depth image information and 2D color image information from the image of the person, and performs image pre-processing; Step 3: S130 obtains the head position and height according to the 3D depth image information. , shoulder position, shoulder area and shoulder width data; step 4, S140 uses 2D color image information for image recognition to obtain an image recognition result; step 5, S150 calculates the gender information according to height, shoulder width and image recognition result Step 6: Based on the top of the head, shoulder position and height, the S160 obtains a body proportion information, and combines the image identification results to identify the age attribute of the person.
由上述流程可進一步得知,步驟一是以透過紅外線式RGB-D感測器安裝於天花板,並以從上而下俯視拍攝取得3D攝影機影像串流;步驟二,是為影像前處理,將3D深度影像資訊與2D彩色影像資訊兩者進行位置對齊校正,再將3D深度影像資訊轉換為灰階影像供後續進行前景物件偵測使用,而2D彩色影像資訊則進行gamma校正,亦供後續使用;步驟三之3D深度影像資訊,是包含頭頂偵測模組,是以3D深度影像透過自適應式高斯混合模型進行背景建模並分離出前景物件,以橢圓頭形偵測找出前景物件頭頂位置,透過與感測器的距離計算出人員身高資訊;以及肩膀偵測模組,是利用頭部位置以輻射狀向外延伸向下搜尋深度值相同之水平平面,經由橢圓形偵測得到長軸之肩膀兩側位置與肩膀區域,利用肩膀兩側位置計算得到肩寬資訊;步驟四之影像辨識,是以一影像辨識模組利用肩膀區域之2D彩色影像以3D深度影像資訊濾除背景影像後,先利用肩膀兩側位置做影像正規化,而後計算其灰階特徵或LBP特徵,並以SVM、Neural Network或AdaBoost之機器學習方法做客層屬性分類,產生各類別機率分數值;步驟五之性別資訊,是以性別辨識模組透過身高與肩寬數值以SVM、Neural Network或AdaBoost之機器學習方法得到性別各類別機率分數後,結合影像辨識機率分數,輸出性別辨識結果;步驟六,是包含頭身比例計算模組,是以頭部到肩膀的距離扣除脖子比例常數得到頭長,再以身高減去頭長得到身長,進而計算出頭身比例資訊;以及年齡辨識模組,是利用頭身比例數值以SVM、Neural Network或AdaBoost之機器學習方法得到年齡各類別機率分數後,結合影像辨識機率分數,輸出年齡辨識結果即得以得到年齡屬性,其年齡屬性, 是分為三類年齡層,包括幼兒,0至12歲、青年,12至18歲,以及成年18歲以上。 It can be further seen from the above process that step one is installed on the ceiling through the infrared RGB-D sensor, and the 3D camera image stream is taken from top to bottom; the second step is for image preprocessing, 3D depth image information and 2D color image information are used for position alignment correction, and then 3D depth image information is converted into grayscale image for subsequent foreground object detection, and 2D color image information is gamma corrected for subsequent use. The 3D depth image information in step 3 includes a head detection module, which is a 3D depth image through an adaptive Gaussian mixture model for background modeling and separation of foreground objects, and an elliptical head detection to find the top of the foreground object. Position, calculate the height information of the person through the distance from the sensor; and the shoulder detection module uses the head position to radially extend downward to search for a horizontal plane of the same depth value, and obtain the length through the elliptical detection The position of the shoulders of the shaft and the shoulder area are calculated by using the position of the shoulders on both sides of the shoulder. The image recognition in step 4 is based on an image. After the module uses the 2D color image of the shoulder region to filter the background image with the 3D depth image information, the image is normalized by using the position on both sides of the shoulder, and then the grayscale feature or the LBP feature is calculated, and the SVM, Neural Network or AdaBoost is used. The machine learning method is used to classify the customer attributes and generate the probability scores of each category. The gender information in step 5 is that the gender identification module obtains the gender probability scores by the machine learning method of SVM, Neural Network or AdaBoost through the height and shoulder width values. After that, combined with the image identification probability score, the gender identification result is output; in step 6, the head body proportion calculation module is included, the head length to the shoulder is subtracted from the neck proportional constant to obtain the head length, and then the height is subtracted from the head length to obtain the body length. And then calculate the ratio information of the head and body; and the age identification module, which uses the machine body learning method of SVM, Neural Network or AdaBoost to obtain the probability scores of various age categories, and combines the image identification probability scores to output the age identification result. To get the age attribute, its age attribute, It is divided into three types of age groups, including young children, 0 to 12 years old, youth, 12 to 18 years old, and adults over 18 years old.
請參閱圖2所示,為本發明垂直式影像人流計數之性別與年齡辨識方法之實施例示意圖,架設於店家天花板之3D攝影機,包含擷取彩色與深度影像資訊,俯視拍攝進入店內之顧客影像,畫面透過USB方式傳送至後端影像辨識則組之主機進行分析與辨識,利用3D深度影像計算人員身高、肩寬與頭身比例資訊,系統即時輸出身高180公分、肩寬40公分與頭身比例1:7之辨識結果;結合2D影像辨識技術,包含辨識肩膀與頭部區域的特徵,以機器學習方法得到人員性別與年齡屬性分析結果,系統即時透過SVM分類器辨識彩色轉灰階之肩膀區域正規化影像,輸出顧客比較偏向男性與成年權重之辨識結果;最後透過綜合3D深度影像特徵如身高、肩寬、頭身比例與影像辨識分數權重,輸出目前進出的顧客為成年男性之屬性特徵。 Please refer to FIG. 2 , which is a schematic diagram of an embodiment of the method for identifying the gender and age of the vertical image flow count of the present invention. The 3D camera mounted on the ceiling of the store includes the color and depth image information, and the customer who enters the store in a bird's eye view. The image is transmitted to the host of the back-end image recognition through the USB mode for analysis and identification. The 3D depth image is used to calculate the height, shoulder width and head-to-head ratio information. The system instantly outputs a height of 180 cm and a shoulder width of 40 cm. The recognition result of the ratio of 1:7; combined with the 2D image recognition technology, including the identification of the characteristics of the shoulder and the head area, the machine learning method to obtain the gender and age attribute analysis results, the system instantly recognizes the color to gray scale through the SVM classifier The shoulder area normalizes the image, and the output customer is biased towards the identification result of male and adult weights. Finally, through the comprehensive 3D depth image features such as height, shoulder width, head body ratio and image recognition score weight, the current customers entering and exiting are the attributes of adult males. feature.
本案技術內容為改善當前人流計數應用,僅能計算人數不能分析其客層屬性之缺失,透過提供多元客層屬性資訊,提升人流計數系統之實用性與產品競爭力。 The technical content of this case is to improve the current flow counting application. Only the number of people can not analyze the lack of customer layer attributes, and improve the practicality and product competitiveness of the flow counting system by providing multiple customer layer attribute information.
上列詳細說明乃針對本發明之一可行實施例進行具體說明,惟該實施例並非用以限制本發明之專利範圍,凡未脫離本發明技藝精神所為之等效實施或變更,均應包含於本案之專利範圍中。 The detailed description of the present invention is intended to be illustrative of a preferred embodiment of the invention, and is not intended to limit the scope of the invention. The patent scope of this case.
綜上所述,本案不僅於技術思想上確屬創新,並具備習用之傳統方法所不及之上述多項功效,已充分符合新穎性及進步性之法定發明專利要件,爰依法提出申請,懇請 貴局核准本件發明專利申請案,以勵發明,至感德便。 To sum up, this case is not only innovative in terms of technical thinking, but also has many of the above-mentioned functions that are not in the traditional methods of the past. It has fully complied with the statutory invention patent requirements of novelty and progressiveness, and applied for it according to law. Approved this invention patent application, in order to invent invention, to the sense of virtue.
S110~S160‧‧‧流程 S110~S160‧‧‧Process
Claims (8)
Priority Applications (2)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
TW106114737A TWI611374B (en) | 2017-05-04 | 2017-05-04 | Gender and age identification method for vertical image flow counting |
CN201710805079.9A CN107590458B (en) | 2017-05-04 | 2017-09-08 | Gender and age identification method of vertical image people flow counting |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
TW106114737A TWI611374B (en) | 2017-05-04 | 2017-05-04 | Gender and age identification method for vertical image flow counting |
Publications (2)
Publication Number | Publication Date |
---|---|
TWI611374B true TWI611374B (en) | 2018-01-11 |
TW201843649A TW201843649A (en) | 2018-12-16 |
Family
ID=61051761
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
TW106114737A TWI611374B (en) | 2017-05-04 | 2017-05-04 | Gender and age identification method for vertical image flow counting |
Country Status (2)
Country | Link |
---|---|
CN (1) | CN107590458B (en) |
TW (1) | TWI611374B (en) |
Cited By (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN110796632A (en) * | 2019-07-30 | 2020-02-14 | 重庆小富农康农业科技服务有限公司 | Pig counting device |
US20210264939A1 (en) * | 2018-06-21 | 2021-08-26 | Nec Corporation | Attribute identifying device, attribute identifying method, and program storage medium |
TWI755960B (en) * | 2020-12-07 | 2022-02-21 | 晶睿通訊股份有限公司 | Object counting method and monitoring camera |
TWI800975B (en) * | 2021-11-09 | 2023-05-01 | 瑞昱半導體股份有限公司 | Image recognition system and training method thereof |
Families Citing this family (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN108830231A (en) * | 2018-06-21 | 2018-11-16 | 芜湖宝文信息科技有限公司 | The statistics equipment of people information statistics for height classification |
CN111474901B (en) * | 2019-12-18 | 2021-04-23 | 山东合众智远信息技术有限公司 | Automated electronic device linkage system and method |
CN111339873B (en) * | 2020-02-18 | 2021-04-20 | 南京甄视智能科技有限公司 | Passenger flow statistical method and device, storage medium and computing equipment |
CN111126357A (en) * | 2020-02-25 | 2020-05-08 | 佛山科学技术学院 | People flow detection device based on image processing |
CN112788818A (en) * | 2020-12-29 | 2021-05-11 | 欧普照明股份有限公司 | Control method, control device and electronic equipment |
Citations (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
WO2012170393A2 (en) * | 2011-06-06 | 2012-12-13 | Aaron Marshall | System and method for providing thermal gender recognition |
TWM443899U (en) * | 2012-02-29 | 2012-12-21 | Bethel Comp Consultant Co Ltd | An intelligent point-of-sales system with active identification of customer group |
Family Cites Families (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN1294521C (en) * | 2004-06-28 | 2007-01-10 | 李剑华 | Outer shape structure of commercial guest greeting robot and identifying method |
CN101587539A (en) * | 2008-05-21 | 2009-11-25 | 上海新联纬讯科技发展有限公司 | Method and system of population flow statistics based on intelligent video identification technology |
CN103186760A (en) * | 2011-12-28 | 2013-07-03 | 昌曜科技股份有限公司 | Pedestrian identification, detection and statistic system |
CN102968738A (en) * | 2012-12-06 | 2013-03-13 | 中国科学院半导体研究所 | Advertising system |
CN105718873B (en) * | 2016-01-18 | 2019-04-19 | 北京联合大学 | Stream of people's analysis method based on binocular vision |
CN106529399A (en) * | 2016-09-26 | 2017-03-22 | 深圳奥比中光科技有限公司 | Human body information acquisition method, device and system |
-
2017
- 2017-05-04 TW TW106114737A patent/TWI611374B/en active
- 2017-09-08 CN CN201710805079.9A patent/CN107590458B/en active Active
Patent Citations (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
WO2012170393A2 (en) * | 2011-06-06 | 2012-12-13 | Aaron Marshall | System and method for providing thermal gender recognition |
TWM443899U (en) * | 2012-02-29 | 2012-12-21 | Bethel Comp Consultant Co Ltd | An intelligent point-of-sales system with active identification of customer group |
Cited By (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20210264939A1 (en) * | 2018-06-21 | 2021-08-26 | Nec Corporation | Attribute identifying device, attribute identifying method, and program storage medium |
CN110796632A (en) * | 2019-07-30 | 2020-02-14 | 重庆小富农康农业科技服务有限公司 | Pig counting device |
CN110796632B (en) * | 2019-07-30 | 2023-08-11 | 重庆渝通合数字科技有限公司 | Pig counting device |
TWI755960B (en) * | 2020-12-07 | 2022-02-21 | 晶睿通訊股份有限公司 | Object counting method and monitoring camera |
US11790657B2 (en) | 2020-12-07 | 2023-10-17 | Vivotek Inc. | Object counting method and surveillance camera |
TWI800975B (en) * | 2021-11-09 | 2023-05-01 | 瑞昱半導體股份有限公司 | Image recognition system and training method thereof |
Also Published As
Publication number | Publication date |
---|---|
TW201843649A (en) | 2018-12-16 |
CN107590458B (en) | 2020-06-23 |
CN107590458A (en) | 2018-01-16 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
TWI611374B (en) | Gender and age identification method for vertical image flow counting | |
JP4198951B2 (en) | Group attribute estimation method and group attribute estimation apparatus | |
CN106241584A (en) | A kind of intelligent video monitoring system based on staircase safety and method | |
CN105138954A (en) | Image automatic screening, query and identification system | |
CN104850825A (en) | Facial image face score calculating method based on convolutional neural network | |
CN106355138A (en) | Face recognition method based on deep learning and key features extraction | |
CN105354527A (en) | Negative expression recognizing and encouraging system | |
Han et al. | Gender classification based on 3D face geometry features using SVM | |
CN104008364A (en) | Face recognition method | |
CN103198330A (en) | Real-time human face attitude estimation method based on depth video streaming | |
CN108647621A (en) | A kind of video analysis processing system and method based on recognition of face | |
Ince et al. | Child and adult classification using ratio of head and body heights in images | |
Putro et al. | Adult image classifiers based on face detection using Viola-Jones method | |
Zhang et al. | Three dimensional binary edge feature representation for pain expression analysis | |
CN115661903B (en) | Picture identification method and device based on space mapping collaborative target filtering | |
Kilinc et al. | Automatic human age estimation using overlapped age groups | |
KR101851823B1 (en) | Method for classifying children and adult bt using head and body height in images | |
Cao et al. | Real-time head detection with kinect for driving fatigue detection | |
Agushinta et al. | Face component extraction using segmentation method on face recognition system | |
Yadappanavar et al. | Machine learning approach for smile detection in real time images | |
Thakur et al. | Identification of face age range group using neural network | |
Thakur et al. | Age identification of facial images using neural network | |
Tabatabaei et al. | Using DLBP texture descriptors and SVM for Down syndrome recognition | |
Mahajan et al. | FCA: A Proposed Method for an Automatic Facial Expression Recognition System using ANN | |
Setthawong et al. | Modified deformable parts model for side profile facial feature detection |