Parking Space Detection
It is a part of the wireless intelligent parking lot management system. A vehicle is attempted to be recognized to judge if corresponding space is occupied.
In our video monitoring cloud platform, a distributed cloud Infrastructure is adopted. At each single cloud, it can achieve more than 640T daily massive video data recording and storage, and support more than 30,000 cameras.
Developed practical across-camera tracking key technology. Innovated the mapping from video space to ground map space. Guarantee non-linear fast video searching based on trajectory visualization and positioning.
Provide resource parallel expansion and plug-in intelligence extension possibilities. Get the most optimal performance by balancing computing resources (CPU, GPU etc.), decoding load and computing demands.
Interactive UI support: Multi-level device management, instant update of various status, switch to various preview layouts, one-key picture capturing, full-functional PTZ controls.
Fruitful recording schedule possibilities: Intuitive time recording strategy, clear inheritance of priorities, selecting and modifying purely by mouse, fine control of start/end recording time.
Flexible personnel management: Institutional organization, permission sets defined as roles, device sets defined as user groups, single person is a combination of certain role and user group.
Reliable maintenance support: Group management of services, scalable system configuration, real-time monitoring of operation data, regular scheduling of tasks and checking of results.
Convenient mobile support: Real-time recording and uploading of video from Android/IOS mobiles (tablets), video remote playback to mobiles.
Featured Add-ons: Distributed service engine, full support of onvif standard, low-latency response, high-throughput storage and network performance.
For intelligent traffic control, accurate vehicle detection is the key. Currently, it is mainly done by the following means: induction loop, video, magnetic detector, ultrasound, radar and infrared. Of all those tools, traditionally the induction loop was used for the most time due its maturity. However, its employment requires a large devastation to the road surface. This would generate damages to the environment, and also need complex later maintenance. A high overall cost is thus expected. And other tools also have their own weaknesses: the ultrasound is easily affected by the wind velocity, the infrared is low in its detection precision, and the radar has a high demand on the installation quality (always a specialist is required). Further, all those 4 tools can only measure limited types of traffic parameters. Therefore, the Video Detector (VD) is winning more and more popularity and support.
It is a part of the wireless intelligent parking lot management system. A vehicle is attempted to be recognized to judge if corresponding space is occupied.
High accurate vehicle detection and license plate recognition are combined, together with in-between automatic 3D PTZ zoom-in operation, to capture illegally parked vehicles 24/7.
Video-based vehicle detectors are installed at every traffic intersection. By detecting and following all vehicles passing by, they can calculate the traffic volume according to a virtual counting gate.
Vehicles are recognized from each frame based on their distinctive feature properties. They are then followed through time by the consistency of their appearances.
Vehicles are recognized from each frame based on their distinctive pairing headlights. The headlights are then followed through time by the geographical continuity and shape consistency.
QanJue Intelligent Box can provide a variety of video processing algorithms for intelligently detection in the input video stream, and can customize the corresponding functional modules according to the needs of the enterprise
QanJue Intelligent Box does not require a professional camera, and can connect to ordinary cameras to realize various intelligent functions such as face recognition, fire detection, off-site detection and fire channel blocking detection. Ordinary cameras’original functions are not affected
The algorithm embedded in QanJue Intelligent Box is highly adaptable and can resist interference in complex environments such as large changes in lighting, complex objects and crowded people, and the recognition accuracy can reach more than 98%.
QanJue Intelligent Box deeply optimizes the real-time identification of dynamic videos, makes intelligent and fully automatic analysis, requires no staff on duty and no manual operation. It issues proactive early warning in case of abnormalities, and communicates such warning via voice, phone call, text message, and other methods.
It can utilize logic computation of time series to intelligently operate results, realize multiplexing of processing output through multi-layer overlay pre-processing and post-processing mechanism, and consider multi-signal fusion and front-end (terminal and cloud) fusion to realize unlimited expansion of intelligent modular assembly.
Real-time automatic monitoring of smoke; real-time automatic monitoring of fire; high-precision active trigger response; automatic alarm system; early warning of fire; multi-module collaboration.
The fire detection module is based on cutting-edge video analysis technology to quickly analyze and identify the smoke and flames in the camera monitoring area. This module has high sensitivity, strong anti-interference ability, and high response speed. It can customize the alarm method for the identified abnormal situations. Not requiring a professional camera, ordinary cameras can also be utilized to achieve intelligent detection.
This module is widely used in large industrial parks, petrochemical warehouses, construction sites, coal mines, logistics warehouses and other places, and is also used in crowded and important security area such as bus stations, railway stations, airports, ports, docks, shopping malls, museums, office parks and workplaces of enterprises and institutions.
Non-cooperative detection of faces; dynamic face recognition; face attribute recognition; face contrast analysis; face identity management; simultaneous recognition of multiple face information
This module uses intelligent face recognition analysis algorithm to accurately detect and identify the facial features of a person. It can detect human faces within one or more camera video frames, and then compare with human face database to provide identity information. It can detect and recognize multiple faces at the same time. It can automatically detect dangerous persons such as wanted criminals and blacklists, and proactively report to the police.
This module is widely used in many important sectors such as public security, transportation, schools, banks, hospitals, customs, military, airports, train stations, etc. It is also used in witness verification, smart access control, attendance management, self-check-in, government affairs hall, security counter-terrorism alarm, exhibition audience verification and other scenes.
Proactive monitoring of attendance; no human supervision required; multi-module collaboration
The main function of this module is to automatically identify the on-the-job situation of staff in the areas under strict monitoring, and to detect and remind them of their behaviors if they leave the post without authorization. This function can strengthen the staff management, improve the responsibilities of staff, avoid loss or accidents caused by leaving their posts, reduce the cost of manpower supervision, and effectively improve management efficiency and safety of the monitored area.
TThis module is widely used in industrial safety production inspection, equipment periodic inspection management, and emergency duty room. It is also used for safety inspection in coal mines, construction sites, riverside and other places.
Real-time and all-round monitoring; simultaneous detection by multiple cameras; fast and automatic judgment and early warning
The main function of this module is to monitor the crowds in public places, detect abnormal behaviors of crowds in a timely manner, and report to the police to take corresponding emergency actions to avoid the occurrence of security incidents such as violent terror, fire, and stomping. This module can monitor in real time whether there are crowds fleeing or being overly agitated in any camera scene. It can automatically judge and alert crowd disturbing behavior caused by dangerous situations such as violence, fire, terrorist attacks, etc., and immediately notify the emergency response department, to prevent the accident from further worsening and minimize human casualties and property losses.
This module is widely used in crowded places such as bus stations, railway stations, airports, parks, squares, shopping centers, schools, and tourist attractions.
Capture of channel blocking and early warning; information push of early warnings; strong environmental adaptability
The main function of this module is to automatically detect the blockage of fire exits such as stairways and aisles. This function is the core in remote monitoring of the smart fire protection platform. Through intelligent detection, event early warning, information push and other functions, it can effectively realize the video surveillance of fire channel blocking, innovate the control and management methods of fire protection through information upgrades, and improve the system of dynamic monitoring, early-warning and forecasting of fire control department.
This module is widely used in fire exits of all buildings and parks, such as residentials, commercial areas, cultural relics and ancient buildings, universities and colleges, health care and industrial parks.
Dynamic non-cooperative automatic statistics; accurate real-time statistics of flow; automatic alarm for passenger flow overload; query and analysis of passenger data
The main function of this module is to automatically count and analyze the flow in the passenger channel through intelligent algorithms. It automatically collects and analyzes data such as the number of people entering or leaving the channel of the designated area, and the flow rate, to form accurate data for decision makers. Dynamic non-cooperative automatic detection can count the incoming and outgoing passenger flow at the same time, and can automatically send alarm for passenger overload.
This module is widely used in shopping centers, schools, hospitals, banks, public security, customs, finance, security and other important industries and fields, as well as the entry places of museums, shopping malls, tourist attractions, exhibitions, chain stores, hotels and other indoor environments.
Allow customized search criteria; allow searching of any area in the video ; allow searching of multiple video files at the same time; efficient and accurate search.
The main function of this module is to use intelligent recognition algorithms to quickly locate and retrieve events, objects and moving objects in different time zones in the video file, and form a list of video clips. The module can quickly and automatically retrieve video files based on search criteria such as the size of the moving object in the screen, the speed of the movement, and the dwell time.
This module can be widely used in public security, transportation, finance, public inspection, shopping malls, new , transportation hubs, crowd gathering places and other industries that involve frequent and time-consuming search of video files.
Non-linear fast positioning and retrieval; automatic tracking of motion trajectory; video area mapping to electronic map; active alarm when entering the restricted area; exclusion recognition of motion interference
This module uses the key technology of cross-camera target tracking independently developed by our company to map the video area space to the electronic map space. The main function is to perform non-linear rapid positioning, retrieval and tracking of the entire network on the real-time to motion trajectory of the calibrated person or object, draw a visual action road map, and proactively alarm if somebody or something enters the restricted area. This module can also read the trajectory of a specific period of its history in the video, obtain the historical information of people or objects, and provide an accurate database for the recording, query, and prediction of motion of object in the video.
This module is widely used in railway stations, bus stations, airports, museums, ports and other key security monitoring areas, as well as unmanned supervision areas such as power stations, oil pipelines, industrial and mining enterprises, railways, tunnels, and key security units, parks and workplaces of government, enterprises and institutions, education industry, etc.
Invisible tags; High recovery rate of tag information; Online watermark addition; Offline watermark traceability; No destroy of the external performance of the video; Tracing the source according to watermark information
The main function of this module is to add invisible but recoverable watermark information to the video that is played back, such as the playback time, the IP address of the playback, the operator, etc. The hidden watermarking information can be used to trace the source of the stolen video, which can effectively alert and prevent the leakage of sensitive information, effectively solve the problem of illegal pirated recording, and play a deterrent role in various illegal pirated recording.
This module is widely used in national security, public security, transportation, radio, television, and video companies, and all other scenarios that need to trace the source of stolen video information, protect video copyright, prevent leakage of internal video and sensitive video.
Automatic identification of data; real-time monitoring of water usage; alarms of abnormal conditions; multi-module collaboration
The function of this module is to realize the automatic identification and reading of the data of various digital meters and pointer dials through intelligent video processing algorithms, and at the same time it can proactively alert for pre-defined abnormal events, such as excess water flow, high or low water level, etc. This module can remotely and automatically read the water meters in real time, and make an analysis and a summary of the data. With one click, an easy-reading report is produced.
This module is widely used in urban overpasses, tunnels, scenic rivers, urban inland rivers, pumping stations, dams, mountain floods and grass-roots flood control, and other scenarios that require the use of water meters.
High accuracy achieved by deep learning; warning of pre-defined behavior; support more expansion of postures.
This module is based on deep learning technology and accurately locates the 18 key nodes of the human body. It accurately recognizes the behavior and posture of the human body in the video, and can provide real-time warning of pre-defined dangerous behaviors and postures, such as fighting, robbery, masking, gathering, crawling, self-harm, falling, etc.
This module is widely used in railway stations, bus stations, airports, tourist attractions and other places where people gather, as well as prisons, detention center and other units that require certain behavior standards. It is also used in elderly homes and communities that care for the elderly.
Vehicle recognition; license plate recognition; high speed of recognition; high accuracy.
The main function of this module is to identify vehicles in the video area and automatically extract vehicle license plate information through intelligent processing algorithm technology. The license plate number is the unique "identity" of the vehicle. The automatic license plate recognition technology can automatically register and verify the "identity" of the vehicle without intrusion. This technology is one important component of the modern intelligent transportation system, and it is essential to maintaining traffic safety and urban security, preventing traffic jams and realizing automatic traffic management.
This module is widely used in highway tolls, parking management, traffic guidance, traffic rule enforcement, highway inspection, vehicle scheduling and other scenarios.
In-house deep learning algorithm
No false detection in 100,000 faces
NNIE NPU acceleration
Strong tolerance against harsh environment
Embedded ID card reader (optional)
Protection level: Robust
Possibility to measure temperature
High reliability and accuracy, fast speed, rich application scenarios, excellent adaptability.
No standing-still is needed, fast passing is therefore guaranteed Embedded ID card reader
Import from clouding platform；
Import from local USB disk；
Import on the unit；
Multiple import possibilities from PC.
Support face liveliness detection;
Not deceived by dolls, pictures and videos etc.
2 USB，Import/export by USB disk;
Highly adaptive to strong light, weak light, back light and side light;
24 hours continuous operation with high accuracy.
Waterproof, dustproof, moisture-proof and anti-riot;
Completely Workable in outside situation.
Support remote device management and data management;
Support MQTT/Http/Https/Socket etc.;
Low power consumption with 24/7 work guarantee.
G01 is suitable for scenarios such as attendance, access control, gate access .
Deep learning core algorithm
10 false recognition based on 50,000 face database
NNIE NPU chipset and Dual-Cortex A7
Self-owned ISP (image signal processing) algorithm adaptable for harsh environment
HiSilicon Chip & Linux OS
Maximum supporting 50,000 face database with 0 false recognition rate
Photo/toy/video are invalid
Millisecond recognition speed, non-stop pass
dustproof, waterproof, moisture-proof, anti-riot
powerful image processing algorithm, all-weather light scene adaptability
MTBF (Mean Time between Failures): 30,000 hours (more than 3 years)
No Sense/ Non-stop dynamic face recognition, fast pass
External USB ID card reader, support for face image and ID card comparison
Supporting multiple face template enrollment mode: Send photo files from Cloud server /Import from U disk/Enroll from terminal / Send from PC tool
Supporting liveness face judgment: photo/doll/video is invalid, to avoid cheating
2 x USB, support for U-disk importing templates/exporting records;
External USB card reader is optional;
Support USB WiFi AP access.
Support for NC, COM, NO
Support Weigand 26/34
Support UART TTL or RS485
Ethernet port: 10/100/1000M adaptable
High quality Sony Sensor with self-owned powerful image processing algorithm, which can ensure to obtain the best quality images in any complex light environment, and ensure face recognition without disturbance all day long.
Support IoT platform for equipment and data management
Support MQTT/Https/Socket communication protocols
SDK supported for the 3rd-party platform integration
Operation temperature: -30゜C, +65゜C
High performance chip, low power consumption, adaptable for high temperature environment
Linux OS, 7 x 24 hours stable operation