Products / Visible face recognition / High Definition Face Analyzer

Active light recognition

super passenger flow

Visible face recognition

RCFRG Barrier-Free Gates

Dynamic Recognition

HD Comparison Ext

Dun Intelligent Access Gate

High Definition Face Analyzer

The High Definition Face Analyzer runs on an X86 platform, and with built-in face recognition algorithms it can independently detect, capture, extract and compare human facial features. Apart from locally-based recognition, the HD Face Analyzer can also access the back-end FACEWIN cloud platform for cloud-based, high-volume face recognition and matching.

一、Functions of the Analyzer:

（1）Face detection: The analyzer is able to detect a human face in a range of environments and determine its position. Its performance is subject to light intensity, background complexity, head tilt angle, and any obscuration.

（2）Feature extraction: Feature information useful for identification is extracted from the detected face image. Identification methods typically include approaches based on geometric features, algebraic features and connection mechanisms.

The extraction of facial feature values is calculated based on a facial image size of CIF/ 4, i.e. 176 × 144, and a quad-core CPU has the capacity to handle 32 images. Obviously, processing capacity will vary in direct proportion to the size of the images, so that with CIF-sized face images (i.e. 352 × 288), the processing limit is 8. The recommended face image size is 176 × 144. Minimum resolution is 80 pixels; in other words, the face in the image must consist of at least 80 pixels.

二、 Technical Features

■ The HD face analyzer comes with built-in face recognition algorithms and can automatically perform dynamic face capture, face filtering, facial feature extraction and feature value matching, and then output the analysis results.

■ Wide applicability – face recognition is possible in a variety of different occasions and light environments (sunlight, normal lighting, etc.).

■ Supports all types of video or image source signals for face matching and analysis, and outputs the recognition results. Both 1 to 1 and 1 to N modes are supported.

■ An alarm function is linked to face authentication results.

■ Standard SDK protocol makes third party calling, access, and output convenient.
三、Product Hardware Specifications

Model：X5000－F1

Embedded architecture, dual-core CPU processor, 1.80 GHz
Storage	64 G capacity, 4 G RAM
Power	Rated voltage and current DC 12V/ 5A
RJ45 interface	1 Ethernet 100/1000M, self-adaptive
USB interface	1 USB 3.0 port
RS485	Optional, no standard
RS232	Optional, no standard
Wireless WiFi	Optional, no standard
GPIO interface	Optional, no standard
LED	Power/ Run, reset switch
Buzzer	Supported
Structural requirements	IP rating 56 or more, can be wall mounted, fully enclosed metal box
Environment/ standards	Operating temperature (-20 to 55 degrees Celsius), humidity 85% RH or less, anti-lightning, anti-surge, runs normally on a 4 KV system and not damaged on 8 KV, CCC, EMC

Model：X5000-F2

CPU	X embedded architecture, quad-core CPU processor，clocked 2.4G
Storage features	Storage type	128 G capacity, 8 G RAM
Hardware Interfaces	Power	Rated voltage and current DC12V/ 5A
	RS485	1, (300-230400 bps)
	RS232	1, (3100-230400 bps)
	RJ45 interface	1, 100/1000M self-adaptive
	GPIO interface	4 input, 4 output
	LED	Power / Run
	Buzzer	Supported
Other features	Environment/ standards	Operating temperature (-20 to 55 degrees Celsius), humidity 85% RH or less, anti-lightning, anti-surge, runs normally on a 4 KV system and not damaged on 8 KV, CCC, EMC