Intelligent video analytics based on AI algorithms and computer vision technology are finding omnipresent applications across industries as they enable real-time monitoring, quick response, and better decision-making. According to the latest research, the global market for intelligent video analytics is expected to reach $26B by 2023. The demand for edge-intelligence based video analytics will increase multifold, driven by emerging applications such as edge computing, autonomous driving, industrial machine vision and medical imaging. e-EYE is a set of services, accelerators, and frameworks derived from our 25+ years of experience to fast-track the development of intelligent, vision-based solutions powered by Edge-AI. e-EYE drastically reduces development efforts and accelerates time-to-market allowing our customers to focus on breakthrough technology innovations.

Why eInfochips for Vision-based Solution?

• Developed 100+ camera vision-based products across different industries such as Automotive, Industrial, Healthcare, Retail etc.
• Designed scalable solutions to manage millions of edge nodes on edge and cloud
• End-to-End video/image pipeline experience (capture, process, record, analyze, stream etc.)
• Expertise around video analytics algorithms development, integration, testing, optimization and porting
• In-house centre of excellence for camera image tuning services Accelerators/IPs – Reusable camera framework, APIs/microservices, RDKs, codecs, stacks, libraries etc.
• In-depth experience across image sensors and optics (Sony, Samsung, Onsemi, OmniVision etc.)

Solution Highlights

Video Management Software

• Multi-tenant federated architecture
• Flexible deployment – on-premise or cloud
• REST APIs, Video-on-Demand, quick ramp-up & ramp-down
• Multi-OS & browser support
• 20+ ready-to-use microservices

Reusable Camera Framework (IP)

• Hardware platform-agnostic, ready to integrate framework
• Plug-n-play, customizable microservices-based architecture
• API support for camera configuration functionalities
• Easy integration with video analytics algorithms
• Low latency, lean memory footprint

Video Analytics Algorithms

• Ready to deploy 22+ typical industry use cases
• Competent team to develop custom algorithms
• Expertise across edge and cloud analytics
• Algorithm optimization/porting on target platforms
• Examples – Object/Pedestrian detection, people counting, face recognition, LPR,
• parking management, Intrusion/Perimeter, Queue management etc.

Multimedia (Audio/Video)

• End-to-end multimedia systems development
• Video pipeline expertise (capture/process/store/transmit, frame stitching, frame rate control)
• Comprehensive video codecs, drivers, interfaces, frameworks,
• SDKs, and protocols support
• Porting & optimization of multimedia frameworks

In-house Image Tuning Lab

• Sensor characterization for major parameters
• Lens and color shading compensation
• Auto exposure and auto white balance tuning
• High-performance, high-quality image pipe tuning
• Subjective and custom image tuning

Accelerators

• Reference designs for IP cameras & NVRs
• RTP/RTSP Client/Server protocol stack
• HTTP live streaming server stack
• 264 HD AVC and SVC encoder
• Video container format (MP4, MPEG2TS)
• Video analytics libraries