【國外新品】2百萬(wàn)像素AI攝像機

　　2-megapixel AI cameras

　　Hanwha Techwin has added five 2-megapixel high-definition AI cameras to its Wisenet P series. The new cameras are designed to complement the 4K models that the manufacturer launched in 2020.

　　As with the 4K models， the new cameras come with Deep Learning AI video analytics which delivers detection accuracy， while minimising false alarms， the developers say. It does so by simultaneously detecting and classifying various object types， such as people， vehicles， faces and car number plates.

　　The AI video analytics is supported by Wisenet AI algorithms which identify the attributes of objects or people， such as their age group and gender or the colour of clothes. The attributes are saved as metadata alongside the images from the Wisenet P series AI cameras. A users can search for specific objects or incidents. The algorithms are even able to recognise if a person is wearing glasses or is holding a bag.

　　Operators can use VMS (video management software) from developers such as Genetec and Milestone Systems， as well as Hanwha Techwin’s Wisenet WAVE and Wisenet SSM， to forensically search for specific objects or people by filtering the recorded metadata， as well as use the search features of the latest Wisenet X network recorders (NVRs) which are able to read AI metadata from the cameras.

　　A recent firmware update， incorporated into the 2-megapixel cameras and available as a downloadable upgrade for the 4K models， sees face mask detection and social distancing analytics added. There is also enhanced people counting and queue management functionality. To help remove the risk of overcrowding and to assist， for example， store managers make best use of staffing， this includes the ability to count the number of people queuing and generate alerts if a set maximum number is passed.

　　The cameras also come with Hanwha Techwin’s Occupancy Monitoring application which generates alarm outputs that can be used to control automatic doors or a traffic beacon， as well as trigger the display of ‘wait’ or ‘enter’ messages when the maximum permitted number of people allowed in an area at one time， has been reached. With these various licence-free Deep Learning AI video analytics applications available as standard on the Wisenet P series AI cameras， users do not have to incur the cost of purchasing licences to be allowed to use them.

　　Features of the new 2-megapixel Wisenet P cameras include：

　　– Digital auto-tracking which enables the cameras to transmit two streams of video， providing the option for operators to see a camera’s full field of view， while at the same time using a digital PTZ auto-tracking feature while viewing Full HD images， to track the movements of a person or vehicle.

　　– WiseNRII optimises image noise reduction to reduce blurring.

　　– 50ma output which can be used to power LEDs， sensors or glass break detectors.

　　The new cameras come with WiseStream III complementary compression. This applies a low compression rate to objects and people which are detected and tracked by AI， while applying high compression to the remaining field of view. When combined with H.265， this results in an improvement in bandwidth efficiency by up to 80 per cent the makers claim compared to H.264.

　　2百萬(wàn)像素AI攝像機

　　Hanwha Techwin公司為其Wisenet P 系列增加了五款2百萬(wàn)像素高清AI攝像機。新款的攝像機旨在作為該公司于2020年推出的4K型號的補充。

　　與4K型號一樣，新相機配備了深度學(xué)習 AI 視頻分析，可提供檢測準確性，同時(shí)最大限度地減少誤報。它通過(guò)同時(shí)檢測和分類(lèi)各種對象類(lèi)型，例如人、車(chē)輛、人臉和車(chē)牌。

　　AI視頻分析由Wisenet AI算法支持，該算法可識別對象或人的屬性，例如年齡組和性別或衣服的顏色。這些屬性與來(lái)Wisenet P系列AI攝相機捕獲的圖像一起保存為元數據，用戶(hù)就可以搜索特定對象或事件。該算法甚至能夠識別一個(gè)人是戴眼鏡還是拿著(zhù)包。

　　運營(yíng)商可以使用Genetec和Milestone Systems等開(kāi)發(fā)商的 VMS(視頻管理軟件)，以及Hanwha Techwin 的Wisenet WAVE和 Wisenet SSM，通過(guò)過(guò)濾記錄的元數據來(lái)取證搜索特定對象或人員，并使用最新的 Wisenet X 網(wǎng)絡(luò )錄像機 (NVR)搜索功能從攝像機讀取AI元數據。

　　最近的固件更新包含在2百萬(wàn)像素攝相機中，可作為4K型號的可下載升級，并且增加了面罩檢測和社交距離分析和增強的人數統計和隊列管理功能。例如，為了幫助消除過(guò)度擁擠的風(fēng)險并協(xié)助商店經(jīng)理充分利用人員配備，這包括能夠計算排隊人數并在超過(guò)設定的最大人數時(shí)發(fā)出警報。

　　這些攝像機還配備了Hanwha Techwin 的占用率監控應用程序，該應用程序可生成警報輸出，可用于控制自動(dòng)門(mén)或交通信標，并在區域內出現允許的最大人數時(shí)觸發(fā)“等待”或“進(jìn)入”的信息。借助Wisenet P系列AI攝像機標配的各種免許可證深度學(xué)習AI視頻分析應用程序，用戶(hù)無(wú)需支付購買(mǎi)許可證的費用即可使用它們。

　　新型2百萬(wàn)像素Wisenet P攝相機的功能包括：

　　– 數字自動(dòng)跟蹤功能，使攝像機能夠傳輸兩個(gè)視頻流，為操作員提供查看攝像機全視野的選項，且在查看全高清圖像的同時(shí)可以使用數字PTZ自動(dòng)跟蹤功能，以便跟蹤人或車(chē)輛的運動(dòng)。

　　– WiseNRII優(yōu)化圖像降噪以減少模糊。

　　– 50ma輸出，可用于為L(cháng)ED、傳感器或玻璃破碎探測器供電。

　　新相機配備了WiseStream III互補壓縮。這對AI 檢測和跟蹤的物體和人物應用低壓縮率，二同時(shí)對其他的視野應用高壓縮率。當與 H.265 結合使用時(shí)，與 H.264 相比，這可將帶寬效率提高多達80%。