| Model | JWXH-401 |
| 4MP AI smart camera | |
| Chip solution | SSC30KQ |
| FLASH/DDR | FLASH: 128MB Memory: 256MB (built-in) |
| System structure | Embedded Linux design, dual-core 32-bit ARM+1T computing power; pure hard compression, watchdog |
| Image sensor | GSC500AI , 1/2.7”, 4 million pixel progressive scan ultra-low illumination CMOS |
| Day and night switching mode | ICR switching, automatic/color/black and white/external trigger |
| Wide Dynamic | Support digital WDR |
| 3D noise reduction | Close/low/medium/high/high |
| Mirror | Close/Horizontal Mirror/Vertical Mirror/180° Rotation |
| Human shape detection | |
| Number of humanoid detections | Can detect ≤5 humanoid targets at the same time |
| Capture response speed | ≤1s |
| Maximum detection frame rate | 6fps/s |
| Detection area | Supports specified area, default full screen |
| Shield area | Supports 1 locale |
| Trigger method | Video trigger |
| Face Detection | |
| Number of face detections | Can detect ≤10 face targets simultaneously |
| Face detection size | Face pixels≥40 pixels |
| Capture response speed | ≤1s |
| Face capture and deduplication | After the face leaves, it will automatically filter and output the optimal face snapshot to reduce repeated snapshots |
| Face scene capture | Support scene image capture and upload |
| Face tracking | Support face detection and tracking |
| Trigger method | Video trigger |
| Electric vehicle detection | |
| Electric vehicle detection size requirements | pixels≧50*50 |
| Actual testing situation | Most electric vehicles of different styles currently on the market can be detected It can be detected when a person is sitting on an electric car, or when a person is pushing an electric car into an elevator |
| Missed detection and false detection | Missed detection: In winter, electric vehicle windshield cloth blocks a large area of electric vehicles, and certain missed detections will occur. False detection: Some strollers with taller wheels and bodies may have certain false detections |
| Face comparison | |
| Contrast type | Supports in-vivo comparison, automatically filters out printed paper, and mobile phone photo/video deception |
| Face database capacity | Supports built-in ≤30,000 faces; supports black and white list control, select 256MB FLASH according to the supported face library capacity) |
| Face comparison recognition rate | 10,000 people database: recognition rate>99% 30,000 people database: recognition rate>98% |
| Face comparison error rate | 10,000 people in the database: misrecognition rate<0.2% 30,000 people in the database: misrecognition rate<0.5% |
| Response speed | ≤200ms |
| Face pose | Supports facial gesture recognition (raise head, lower head, turn left, turn right) |
| expression | Slight expressions do not affect facial comparison |
| Human Structure | |
| Facial features | Supports age prediction, gender prediction, glasses, beard, and smile detection |
| Mask detection | Support mask detection |
| Hard hat detection | Supports helmet detection |
| Stranger identification | Supports detection and recognition of stranger faces |
| Car structured | |
| License plate detection | Detect whether there is a license plate |
| License plate detection size | ≥40*80 |
| License plate number recognition | Support |
| License plate color recognition | Blue license plate, green license plate, yellow license plate, white license plate, yellow-green license plate |
| License plate type recognition | Civilian vehicles, military vehicles, police vehicles |
| License plate library capacity | Supports built-in license plate number ≤30,000; supports black and white list control |
| Vehicle type recognition | Trucks, cars, |
| The impact of car lights | The vehicle turns on the high beam, but the low beam has no effect on license plate detection |
| Capture response speed | ≤1s |
| Vehicle color detection | Support |
| Audio and video encoding | |
| Video encoding standard | H.265/H.264 |
| Encoding characteristics | Main stream: 2560*1440@25fps / 1920*1080@25fps / 1280*720@25fps Sub-stream: 1280*720@15fps / 960*540@15fps / 640*360@25fps |
| Maximum image resolution | 2560*1440 |
| Video bitrate | Fixed code rate, variable code rate (192kbps~8000kbps) |
| Video type | Composite stream/video stream |
| Audio coding standard | G.711A / AAC 64Kpbs |
| Network features | |
| Network Protocol | TCP/IP, HTTP, DHCP, RTSP, NTP, P2P |
| Interface protocol | ONVIF, GB28181/2016, GAT-1400 Sichuan Telecom-Magic Mirror |
| Number of simultaneous preview videos | 2 |
| Safe mode | Authorized username and password, RTSP verification |
| Motion detection | Set detection area, multi-level sensitivity can be set |
| Event linkage | Capture, audio linkage, alarm light linkage |
| Character overlay (OSD) | Title, time and date overlay, OSD supports 7 groups of customized OSD |
| System upgrade | Support tool upgrade |
| Interface | |
| Network interface | RJ45 10M/100M adaptive Ethernet port |
| Audio | Built-in MIC, built-in power amplifier + external speaker |
| Power interface | POE network cable power supply / 1 DC 12V power interface |
| Reset | Support hardware reset |
| Operating temperature/humidity | -40℃~+80℃ / ≤90% |
| Wi-Fi | |
| Frequency band | 2.4G specification 802.11b/g/n/ax |
| Docking module | a69c:88dc |