Reduce the intensity of comparison work and improve audit efficiency
Bill is the core of finance. STR intelligent bill recognition is the key to realize the financial sharing center. Intelligent bill recognition provides reliable electronic image data service for invoice reimbursement of enterprises and institutions, and provides electronic bill interface for self-service document delivery, online approval, cashier payment, business audit and financial audit.
Provide bill "data structure" based on STR artificial intelligence technologies such as robot learning, natural language processing, intelligent correction, etc., improve the work efficiency of financial personnel, reduce the labor intensity of repetitive work, provide more scientific data support for business system formulation, reduce business risks, and truly empower the development of enterprises.
Mobile APP bill collection can provide the possibility for efficient financial operation of enterprises, and improve the efficiency of financial operation without increasing the financial burden; Intelligent bill recognition provides reliable image bills and structured data for finance, quickly establishes orderly bill management for enterprises, and easily realizes true electronic vouchers. At the same time, intelligent screening, intelligent sorting, classified storage, real-time status monitoring, security and tamper-proof are carried out for massive bills.
The LeoStor distributed storage system provides powerful parallel read and write support for business systems. Its fast retrieval of tens of millions of small files, millisecond file read and write delay, single volume up to 300PB, and ServerRaid and DiskRaid data protection can enable enterprises to use it without worry for more than 30 years, making financial data safer and more reliable; The object storage technology provided by LeoStor provides multi-platform unified data interface, and can easily realize multi-financial center bill sharing and docking blockchain business.
The system has the following advantages
- Reliability: Cloud computing hyper-integrated cluster elastic architecture design is adopted, and the reliability exceeds 99.999%;
- Security: Use ticket fingerprint, blockchain, WORM tamper-proof storage technology and ACL access control technology to make data more secure;
- Progressiveness: LoongShine STR technology can identify scenes with blur, skew, reflection, deformation, long distance and low pixels, and multi ticket integration;
- Efficiency: Second-level fast identification and fast docking. A single intelligent accounting robot accounts for 2000 bills per hour. Through clustering processing, it can process millions of bills per day;
- Accuracy: The traditional OCR text recognition rate can only reach 40%, while the LoongShine STR intelligent bill recognition rate exceeds 95%. Moreover, the algorithm can continue to train itself, and the recognition rate can still be further improved;
- Multi-channel collection: bill entry can be carried out through mobile phone APP, high-resolution camera, scanner, camera and other devices to meet the needs of various use scenarios;
- Deliverable system: including LoongShine mobile ticket collection subsystem, LoongShine STR intelligent identification subsystem,Kirin balance flow balance subsystemand LeoStor distributed storage subsystem, one-stop delivery
- Mass storage:adopt distributed storage technology to meet the storage needs of large enterprises and institutions for 50 years;
It supports the identification of VAT invoices, train tickets, air transport electronic ticket itinerary tickets, taxi tickets, roll tickets, quota invoices, bridge toll invoices and other full ticket types, and provides the interface between the RestFul SDK API and financial accounting software such as Kingdee and UFIDA,to build a new generation of intelligent enterprise financial system centered on original vouchers, supplemented by business, projects, and organizational structures,to reduce human intervention, and realize the automatic processing of enterprise finance and tax, It is an essential core system for building the next generation of enterprise and public finance system.
Adopt AI and big data analysis technology to provide more efficient auxiliary means for enterprise internal audit, listing audit and government audit, improve audit efficiency, shorten audit cycle, and truly reduce the burden of audit.
Multifocal segment intelligent fusion splicing 500 million pixels
In traditional video surveillance, to cover an area without blind spots, it is necessary to deploy cameras with multiple focus segments, or deploy a camera stand at each distance. This not only increases the cost of infrastructure, but also on the large monitoring screen, if you want to see the whole situation, you can only make each monitoring screen correspond to one camera. The whole monitoring screen is messy. Intelligent fusion splicing solves these two problems, Through the integrated multi-camera and AI fusion splicing technology, the implementation is simpler and the monitoring screen is clearer and smoother, while the multi-focus real-time splicing fusion technology has the leading level in the industry.
Intelligent panoramic imaging technology is to integrate multiple photoelectric imaging systems and cooperate with multi-channel electronic processing systems to achieve imaging of hundreds of millions or billions of pixels. This technology can solve the contradiction between the large field of view and high resolution of the traditional photoelectric system, greatly improve the total amount of information transmitted by the photoelectric system, and promote the development of the photoelectric system of big data information towards a unified platform integrating image, communication, storage, intelligent analysis and application;
In order to achieve smoother splicing and access effects, the system has solved many difficult problems, such as:
- Solve the problem of image alignment due to dislocation between images;
- Solve the problem of image synthesis of color difference between image pixels;
- High-speed video splicing;
- Detection and matching of image feature points;
- Ghost elimination problem;
It supports multiple sets of high-definition cameras and intelligent video splicing technology, and can monitor the following areas in real time on the original large monitoring screen:
- Large regional class (A series): plaza, airport, venue, stadium, station, scenic spot, power plant, sky, forest, farmland, aerial reconnaissance;
- Horizontal path class (W series): river, airport runway, park boundary, border defense, coastal defense;
- Longitudinal path class (L series): municipal roads, highways, railways, tunnels, bridges;
Adopt virtual camera technology, support HTTP protocol, ONVIF protocol, GB-28181 standard and other methods to access virtual camera, do not change the topology of the user's existing monitoring platform, and seamlessly access the existing monitoring platform based on the existing network; It supports H264 and H265 video coding formats, has its own video structured system, and can also seamlessly connect with third-party video structured system; At the same time, the system provides a variety of video codec and multi-stream solutions for mobile phones, Pads, laptops, desktops and other devices to browse.
Provide delivery system, including panoramic camera system, AI intelligent splicing system, video monitoring system, encoding and decoding and transcoding system, AI intelligent identification system, LeoStor parallel storage system, LDC1000 modular cabinet system, 10 Gigabit network equipment, one-stop delivery, operation and maintenance.
