TWI723383B - Image recognition device and method - Google Patents
Image recognition device and method Download PDFInfo
- Publication number
- TWI723383B TWI723383B TW108114041A TW108114041A TWI723383B TW I723383 B TWI723383 B TW I723383B TW 108114041 A TW108114041 A TW 108114041A TW 108114041 A TW108114041 A TW 108114041A TW I723383 B TWI723383 B TW I723383B
- Authority
- TW
- Taiwan
- Prior art keywords
- image
- processing module
- frame
- image frame
- video stream
- Prior art date
Links
Images
Landscapes
- Image Analysis (AREA)
Abstract
Description
本發明涉及影像處理技術領域,尤其涉及一種圖像識別裝置及方法。 The present invention relates to the field of image processing technology, in particular to an image recognition device and method.
隨著高清顯示技術之發展,市面上出現了4K(圖元解析度4096*2160)之視頻播放設備,未來甚至可上市8K(圖元解析度7680*4320)之視頻播放設備。該些視頻播放設備之幀頻一般為30FPS,或者更高。當需要對每幀圖像進行一個或多個目標物檢測時,由於視頻幀之圖元較高且幀頻較大,導致圖像識別工作量較大且識別速度要求要快,習知之圖像識別模組之處理器處理速度有限,無法做到對每一播放之超高清幀圖像進行即時目標物檢測。 With the development of high-definition display technology, 4K (picture element resolution 4096*2160) video playback equipment appeared on the market, and 8K (picture element resolution 7680*4320) video playback equipment may even be marketed in the future. The frame rate of these video playback devices is generally 30FPS or higher. When it is necessary to detect one or more target objects in each frame of image, due to the higher image elements of the video frame and the larger frame rate, the image recognition workload is larger and the recognition speed is required to be fast. The conventional image The processing speed of the processor of the recognition module is limited, and it is unable to perform real-time target detection for each UHD frame image played.
有鑑於此,有必要提供一種圖像識別裝置及方法,能夠提高圖像識別效率,實現對每一超高清幀圖像進行即時目標物檢測。 In view of this, it is necessary to provide an image recognition device and method that can improve the efficiency of image recognition and realize real-time target detection for each ultra-high-definition frame image.
本發明一實施方式提供一種圖像識別方法,應用於圖像識別裝置,所述圖像識別裝置包括第一處理模組、存儲模組、第二處理模組,所述方法包括:利用所述第一處理模組獲取待處理之視頻流,並對所述視頻流進行預處理得到按幀播放順序排列之圖像佇列;將所述圖像佇列存儲至所述存儲模組;及利用所述第二處理模組從所述存儲模組中讀取待識別之圖像幀,並對所述圖像 幀進行識別,以檢測所述圖像幀中之至少一目標物;其中,所述第二處理模組根據所述圖像佇列之圖像幀排序進行依次讀取。 An embodiment of the present invention provides an image recognition method applied to an image recognition device. The image recognition device includes a first processing module, a storage module, and a second processing module. The method includes: using the The first processing module obtains the video stream to be processed, and preprocesses the video stream to obtain an image queue arranged in the order of frame playback; stores the image queue in the storage module; and uses The second processing module reads the image frame to be recognized from the storage module, and compares the image The frame is recognized to detect at least one target in the image frame; wherein, the second processing module reads sequentially according to the sequence of the image frames in the image queue.
本發明一實施方式還提供一種圖像識別裝置,包括第一處理模組、存儲模組、第二處理模組及通信匯流排,所述第一處理模組、所述存儲模組及所述第二處理模組藉由所述通信匯流排進行相互間之通信;所述第一處理模組用於獲取待處理之視頻流,並對所述視頻流進行預處理得到按幀播放順序排列之圖像佇列;所述第一處理模組還用於將所述圖像佇列存儲至所述存儲模組;所述第二處理模組用於從所述存儲模組中讀取待識別之圖像幀,並對所述圖像幀進行識別,以檢測所述圖像幀中之至少一目標物;其中,所述第二處理模組根據所述圖像佇列之圖像幀排序進行依次讀取。 An embodiment of the present invention also provides an image recognition device, including a first processing module, a storage module, a second processing module, and a communication bus. The first processing module, the storage module, and the The second processing module communicates with each other through the communication bus; the first processing module is used to obtain the to-be-processed video stream, and preprocess the video stream to obtain the order of frame playback Image queue; the first processing module is also used to store the image queue to the storage module; the second processing module is used to read the to-be-identified from the storage module To identify the image frame to detect at least one target in the image frame; wherein, the second processing module sorts the image frames according to the image queue Read sequentially.
與習知技術相比,上述圖像識別裝置及方法,利用第一處理模組進行視頻流解碼得到待識別之圖像幀,利用第二處理模組之多執行緒對圖像幀進行識別,從而提高圖像識別效率,實現對每一超高清幀圖像進行即時目標物檢測。 Compared with the prior art, the above-mentioned image recognition device and method use the first processing module to decode the video stream to obtain the image frame to be recognized, and use the multi-thread of the second processing module to recognize the image frame. Thereby improving the efficiency of image recognition and real-time target detection for each ultra-high-definition frame image.
10:第一處理模組 10: The first processing module
20:存儲模組 20: Storage module
30:第二處理模組 30: The second processing module
40:通信匯流排 40: Communication bus
100:圖像識別裝置 100: Image recognition device
P#1~P#N:執行緒
PU#1~PU#N:處理器單元
圖1是本發明一實施方式之圖像識別裝置之架構示意圖。 FIG. 1 is a schematic diagram of the structure of an image recognition device according to an embodiment of the present invention.
圖2是本發明一實施方式之圖像識別裝置之功能模組圖。 Fig. 2 is a functional block diagram of an image recognition device according to an embodiment of the present invention.
圖3是本發明一實施方式之圖像識別方法之流程圖。 Fig. 3 is a flowchart of an image recognition method according to an embodiment of the present invention.
請參閱圖1,為本發明圖像識別裝置一較佳實施例之示意圖。 Please refer to FIG. 1, which is a schematic diagram of a preferred embodiment of the image recognition device of the present invention.
所述圖像識別裝置100包括第一處理模組10、存儲模組20、第二處理模組30及通信匯流排40。所述第一處理模組10、所述存儲模組20及所述
第二處理模組30藉由所述通信匯流排40來完成相互間之通信。
The
所述存儲模組20可包括至少一記憶體。所述記憶體可包括高速隨機存取記憶體,還可包括非易失性記憶體,例如硬碟機、記憶體、插接式硬碟機,智慧存儲卡(Smart Media Card,SMC),安全數位(Secure Digital,SD)卡,快閃記憶體卡(Flash Card)、至少一個磁碟記憶體件、快閃記憶體器件、或其他易失性固態記憶體件。所述通信匯流排40包括但不限於資料匯流排、電源匯流排、控制匯流排、狀態信號匯流排等。
The
所述第一處理模組10用於獲取待處理之視頻流,並對所述視頻流進行預處理得到按幀播放順序排列之圖像佇列。
The
於一實施方式中,所述第一處理模組10可從具有視頻錄製功能之設備、視頻播放設備、存放裝置等中獲取待處理之視頻流。所述預處理可包括視頻流之解碼與幀圖像分割處理,幀圖像分割處理可是將一視頻分割成多個圖像幀。可理解視頻播放原理即快速連續地播放圖像幀,所述第一處理模組10優選將所述視頻流處理成按幀播放順序排列之圖像佇列,該圖像佇列之起始圖像幀可是該視頻流之第一幀圖像,該圖像佇列之末端圖像幀可是該視頻流之最後一幀圖像。
In one embodiment, the
為便於後續對圖像幀之識別工作,所述第一處理模組10還用於將所述圖像佇列存儲至所述存儲模組20。
In order to facilitate subsequent identification of image frames, the
所述待處理之視頻流可是存儲於存儲模組20中之視頻流,進而所述第一處理模組10可從所述存儲模組20中獲取所述待處理之視頻流。
The video stream to be processed may be a video stream stored in the
於一實施方式中,所述第一處理模組10可包括一個或多個處理器,所述處理器可是中央處理單元(Central Processing Unit,CPU),微處理器、數位訊號處理器(Digital Signal Processor,DSP)、專用積體電路(Application Specific Integrated Circuit,ASIC)、現成可程式設計閘陣列(Field-Programmable Gate Array,
FPGA)或者其他可程式設計邏輯器件、分立門或者電晶體邏輯器件、分立硬體元件等。當所述第一處理模組10包括多個處理器時,所述第一處理模組10可採用多執行緒來對所述視頻流進行預處理得到按幀播放順序排列之圖像佇列。
In one embodiment, the
所述第二處理模組30用於從所述存儲模組20中讀取待識別之圖像幀,並對所述圖像幀進行識別,以檢測所述圖像幀中之至少一目標物。
The
於一實施方式中,所述待識別之圖像幀可是所述圖像佇列中未進行圖像識別之圖像幀。所述第二處理模組30可包括至少一個圖像識別模型,以對所述圖像幀進行識別。所述圖像識別模型可是藉由機器學習演算法訓練得到之識別模型,比如深度殘差網路演算法、卷積神經網路演算法等。
In one embodiment, the image frame to be recognized may be an image frame in the image queue for which no image recognition has been performed. The
所述目標物可根據實際使用需求進行設定,比如所述目標物可是人、包(手提包、背包)等。當所述第二處理模組30需要識別多個類型之目標時,每一類型之目標物可對應一圖像識別模型。比如所述第二處理模組30包括有第一圖像識別模型及第二圖像識別模型,所述第一圖像識別模型用於識別圖像幀中之人,所述第二圖像識別模型用於識別圖像幀中之包,則所述第二處理模組30可實現檢測圖像幀中之人與包。
The target can be set according to actual usage requirements. For example, the target can be a person, a bag (handbag, backpack), and the like. When the
於一實施方式中,所述第二處理模組30優選根據所述圖像佇列之圖像幀排序進行依次讀取與識別,即圖像佇列中排序靠前之圖像幀會優先被所述第二處理模組30讀取與識別。
In one embodiment, the
於一實施方式中,所述第二處理模組30從所述存儲模組20中讀取到圖像幀後,還對所述圖像幀進行參數調整,然後再對參數調整後之圖像幀進行識別,其中所述參數調整可至少包括圖元調整與/或亮度調整,進而可提高圖像識別之效率與準確度。所述圖元調整比如可是將圖像幀之圖元縮小以減少計算量。
In one embodiment, after the
於一實施方式中,所述第二處理模組30還用於將識別得到之所述
目標物之物體類別資訊作為所述目標物之標籤資訊儲存至所述圖像幀。
In one embodiment, the
請同時參閱圖2,所述第二處理模組30可是一處理器集群,所述處理器集群支援硬體擴展。所述處理器集群包括多個執行緒P#1~P#N及多個處理器單元PU#1~PU#N,N為大於1之自然數。每一執行緒一一對應每一處理器單元。比如,所述執行緒P#1用於每次從所述存儲模組20中讀取一幀待識別之圖像並傳送至所述處理器單元PU#1,以對所述至少一目標物進行識別。每一所述處理器單元PU#1~PU#N均包括至少一圖像識別模型。所述第二處理模組30藉由多執行緒P#1~P#N與多處理器單元PU#1~PU#N之配合可極大提高圖像識別速度,實現對超高清圖像幀進行即時目標物檢測。
Please also refer to FIG. 2, the
於一實施方式中,由於超高清圖像幀之資料量較大,所述執行緒P#1每次從所述存儲模組20中讀取一幀待識別之圖像,避免資料量讀取過大影響識別速度,然後傳送給處理器單元PU#1進行識別,當處理器單元PU#1識別完成一幀圖像後,執行緒P#1再讀取下一幀待識別圖像。
In one embodiment, due to the large amount of data in the ultra-high-definition image frame, the
可理解之是,當所述存儲模組20存儲有未識別之圖像幀時,多個執行緒程P#1~P#N會並行去讀取所述未識別之圖像幀。比如所述處理器集群包括八個執行緒P#1~P#8及八個處理器單元PU#1~PU#8,執行緒P#1讀取所述圖像佇列之第一圖像幀並傳送給處理器單元PU#1進行識別,執行緒P#2讀取所述圖像佇列之第二圖像幀並傳送給處理器單元PU#2進行識別,執行緒P#8讀取所述圖像佇列之第八圖像幀並傳送給處理器單元PU#8進行識別。當處理器單元PU#1完成第一圖像幀之識別後,執行緒P#1會再次讀取所述圖像佇列之第九圖像幀並傳送給處理器單元PU#1進行識別。
It is understandable that when the
於一實施方式中,所述處理器單元PU#1~PU#N可是中央處理單元(Central Processing Unit,CPU),微處理器、數位訊號處理器(Digital Signal Processor,DSP)、專用積體電路(Application Specific Integrated Circuit,ASIC)、
現成可程式設計閘陣列(Field-Programmable Gate Array,FPGA)或者其他可程式設計邏輯器件、分立門或者電晶體邏輯器件、分立硬體元件等。
In one embodiment, the processor
圖3為本發明一實施方式中圖像識別方法之流程圖。根據不同之需求,所述流程圖中步驟之順序可改變,某些步驟可省略。 Fig. 3 is a flowchart of an image recognition method in an embodiment of the present invention. According to different needs, the order of the steps in the flowchart can be changed, and some steps can be omitted.
步驟S300,利用所述第一處理模組10獲取待處理之視頻流,並對所述視頻流進行預處理得到按幀播放順序排列之圖像佇列。
In step S300, the
步驟S302,將所述圖像佇列存儲至所述存儲模組20。
Step S302, storing the image queue in the
步驟S304,利用所述第二處理模組30從所述存儲模組20中讀取待識別之圖像幀,並對所述圖像幀進行識別,以檢測所述圖像幀中之至少一目標物。
Step S304: Use the
於一實施方式中,所述第二處理模組30優選根據所述圖像佇列之圖像幀排序進行依次讀取。
In one embodiment, the
上述圖像識別裝置及方法,利用第一處理模組進行視頻流解碼得到待識別之圖像幀,利用第二處理模組之多執行緒對圖像幀進行識別,從而提高圖像識別效率,實現對每一超高清幀圖像進行即時目標物檢測。 The above-mentioned image recognition device and method use the first processing module to decode the video stream to obtain the image frame to be recognized, and use the multiple threads of the second processing module to recognize the image frame, thereby improving the efficiency of image recognition. Realize real-time target detection for each ultra-high-definition frame image.
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,本發明之範圍並不以上述實施方式為限,舉凡熟悉本案技藝之人士爰依本發明之精神所作之等效修飾或變化,皆應涵蓋於以下申請專利範圍內。 In summary, the present invention meets the requirements of an invention patent, and Yan filed a patent application in accordance with the law. However, the above are only the preferred embodiments of the present invention, and the scope of the present invention is not limited to the above embodiments. Anyone familiar with the art of the present case makes equivalent modifications or changes based on the spirit of the present invention. Should be covered in the scope of the following patent applications.
Claims (10)
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
TW108114041A TWI723383B (en) | 2019-04-22 | 2019-04-22 | Image recognition device and method |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
TW108114041A TWI723383B (en) | 2019-04-22 | 2019-04-22 | Image recognition device and method |
Publications (2)
Publication Number | Publication Date |
---|---|
TW202040424A TW202040424A (en) | 2020-11-01 |
TWI723383B true TWI723383B (en) | 2021-04-01 |
Family
ID=74201439
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
TW108114041A TWI723383B (en) | 2019-04-22 | 2019-04-22 | Image recognition device and method |
Country Status (1)
Country | Link |
---|---|
TW (1) | TWI723383B (en) |
Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20060209957A1 (en) * | 2002-11-26 | 2006-09-21 | Koninklijke Philips Electronics N.V. | Motion sequence pattern detection |
US9055176B1 (en) * | 2012-11-16 | 2015-06-09 | Altera Corporation | Method and apparatus for performing multi-cadence detection in a motion adaptive video deinterlacer |
CN105264879A (en) * | 2013-06-14 | 2016-01-20 | 高通股份有限公司 | Computer vision application processing |
CN108028902A (en) * | 2015-07-16 | 2018-05-11 | 博拉斯特运动有限公司 | Integrated sensor and video motion analysis method |
-
2019
- 2019-04-22 TW TW108114041A patent/TWI723383B/en active
Patent Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20060209957A1 (en) * | 2002-11-26 | 2006-09-21 | Koninklijke Philips Electronics N.V. | Motion sequence pattern detection |
US9055176B1 (en) * | 2012-11-16 | 2015-06-09 | Altera Corporation | Method and apparatus for performing multi-cadence detection in a motion adaptive video deinterlacer |
CN105264879A (en) * | 2013-06-14 | 2016-01-20 | 高通股份有限公司 | Computer vision application processing |
CN108028902A (en) * | 2015-07-16 | 2018-05-11 | 博拉斯特运动有限公司 | Integrated sensor and video motion analysis method |
Also Published As
Publication number | Publication date |
---|---|
TW202040424A (en) | 2020-11-01 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN111832366B (en) | Image recognition apparatus and method | |
Varol et al. | Long-term temporal convolutions for action recognition | |
US10599958B2 (en) | Method and system for classifying an object-of-interest using an artificial neural network | |
US11830202B2 (en) | System and method for generating player tracking data from broadcast video | |
CN106687989B (en) | Method, system, readable medium and apparatus for facial expression recognition | |
US11875558B2 (en) | Methods and apparatus to generate temporal representations for action recognition systems | |
EP4035070B1 (en) | Method and server for facilitating improved training of a supervised machine learning process | |
CN111552837A (en) | Animal video tag automatic generation method based on deep learning, terminal and medium | |
TW202024992A (en) | Comparison method and device based on multiple frames of face images and electronic equipment | |
Yin et al. | A face anti-spoofing method based on optical flow field | |
US20200074638A1 (en) | Image segmentation method, apparatus and non-transitory computer readable medium of the same | |
TW202314634A (en) | Image processing method, electronic device and computer readable storage medium | |
KR102421604B1 (en) | Image processing methods, devices and electronic devices | |
CN106874922B (en) | Method and device for determining service parameters | |
TWI584227B (en) | Image processing method, image processing device and image processing system | |
TWI723383B (en) | Image recognition device and method | |
CN110363187B (en) | Face recognition method, face recognition device, machine readable medium and equipment | |
Yanakova et al. | Facial recognition technology on ELcore semantic processors for smart cameras | |
Selvi et al. | FPGA implementation of a face recognition system | |
CN113591656A (en) | Image processing method, system, device, equipment and computer storage medium | |
Lai et al. | Communication-aware face detection using noc architecture | |
Charran et al. | Real-Time Identity Censorship of Videos to Enable Live Telecast Using NVIDIA Jetson Nano | |
CN110210306B (en) | Face tracking method and camera | |
CN107665341A (en) | One kind identification control method, electronic equipment and computer product | |
Chen et al. | An Enhanced LBPH Approach to Ambient-Light-Affected Face Recognition Data in Sensor Network. Electronics 2023, 12, 166 |