This webinar discusses compression, an often misunderstood, undervalued and under-deployed tool for vision applications. This session will discuss the various approaches to deploying compression, and how a novel approach enables low latency, high reliability, and lossless compression for vision applications. When designed and deployed properly compression can significantly boost the bandwidth capabilities of existing infrastructure and increase performance while maintaining and lowering costs.
Overview
This webinar discusses compression, an often misunderstood, undervalued and under-deployed tool for vision applications. This session will discuss the various approaches to deploying compression, and how a novel approach enables low latency, high reliability, and lossless compression for vision applications.
When designed and deployed properly compression can significantly boost the bandwidth capabilities of existing infrastructure and increase performance while maintaining and lowering costs.
With Hosts James Falconer, Product Manager & Eric Boisvert, Principal Hardware Designer.
The session will include real-world application examples of compression deployed in high-value imaging systems for security, defence, and medical devices, such as:
- Dental / Medical Imaging: This requires larger image sensors and more data but without the need to invest and replace installed infrastructure. For patients, the benefit is reducing exposure times and not requiring second scans.
- 360 Imaging / Machine Vision: This empowers more image sources, 24/7 monitoring, ability to support larger image sensors and multiple networked devices over lower-cost infrastructure.
Key Takeaways
When properly deployed, Compression empowers
- Higher throughput on existing networks
- Cost savings by avoiding unnecessary infrastructure upgrades
- Adds Minimal induced low latency for real-time applications
Growing Data Demands and The Need for Higher Resolution and Frame Rates
- Medical imaging - larger image sensors such as dental and medical X-rays, generate larger data files
- Machine Vision - 24/7 monitoring requires real-time processing of multiple high-resolution feeds
- In these instances, the main challenge is data bottlenecks in existing Ethernet infrastructure
Bandwidth Limitations & Infrastructure Costs
- Compression Trade-offs - lossy methods sacrifice image integrity; uncompressed data overloads networks
- Upgrading Infrastructure is Expensive - many industries operate on legacy networks that are not designed for today’s data loads
Performance & Latency Concerns
- Medical Imaging - as data increases, potential transmission delays result in errors that impact processing and analysis, requiring additional scans that add procedure costs
- Machine Vision - real-time quality control and security systems require instant image transmission
Here are some emerging trends that will benefit from advanced compression
- Multi-Sensor & 3D Vision - expanding beyond 2D imaging requires even more efficient data handling, multiple sensors and fusion of data from multiple sensors also requires more efficient data transfer for industries that need to maintain existing infrastructure.
- Cloud-Based Imaging Platforms - telemedicine & remote monitoring demand fast, reliable compression
- AI & Deep Learning - high-quality data is essential for AI-driven diagnostics and automated defect / object detection
Speakers
James Falconer has a solid background in engineering and aerospace, with experience in research, teaching, and product management. Currently, James is Product Manager at Pleora Technologies and serves as the Vice-Chair of the GigE Vision Technical Committee at AIA - Advancing Vision + Imaging. Previous roles include being a Senior Field Application Engineer and Production Engineer. James has a Master's degree in Applied Science in Aerospace Engineering from Carleton University.
Eric Boisvert has played a pivotal role in Pleora success, having been a part of the company since its inception in 2000. He is currently a Principal Hardware Designer with over 30 years of extensive expertise in FPGA and RTL design. His focuses also include PCB design, schematic entry and debugging, as well as machine vision and image processing. Eric has an impressive background with 3 patents, and a Bachelor of Electrical Engineering from Laval University.