Home

Solutions

Technology Trends

There is a significant trend towards embedding video analytics on edge-of-network devices. Some vendors are embedding analytics on digital signal processors (DSP) with processing capabilities and working in conjunction with IP camera and encoder manufacturers to provide intelligent edge devices. This has several theoretical benefits; however the current state of embedded technology presents some significant obstacles.

Pros

Decreased Bandwidth and Storage requirements = Significant Cost Savings
Embedding analytics on edge-of-network devices allows the processing of the video feed to occur on the edge. End-users can then choose only to send significant data to the server. This decreases both bandwidth and storage requirements, which can result in significant cost savings.

Reduced Acquisition & Maintenance
Embedding analytics on edge of network devices means greatly reduces server hardware requirements, significantly reduing acquisition costs. Additionally, with less hardware, maintenance requirements are lessened, as are footprint requirements, further lowering the total cost of ownership.

Cons

Limited Processing Power = Reduced Accuracy
The major drawback of analytics on the edge is the limited processing power of cost-effective DSPs. Many high quality video analytics algorithms have a large footprint and most DSPs do not have the memory and processing power to run them efficiently. Reduced processing power reduces the accuracy of the analytics which can lead to an increase in both false positive and false negative alerts. In a market that is demanding higher levels of accuracy from video analytics, this presents a significant problem.

Labor Intensive Configuration
Another drawback of embedding analytics in encoders, camera hardware, or DSPs is the fact that each device needs to be configured individually. This increases installation time and often results in unexpected costs.