Natural Scene Derived Camera Edge Spatial Frequency Response for Autonomous Vision Systems

The edge-based Spatial Frequency Response (e-SFR) is an established measure for camera system quality performance, traditionally measured under laboratory conditions. With the increasing use of Deep Neural Networks (DNNs) in autonomous vision systems, input signal becomes crucial optimal operation. This paper proposes a method to estimate e-SFR from pictorial natural scene derived SFRs (NSSFRs) as previously presented, laying foundation adapting traditional real-time measure.In this study, NS-SFR parameter variations are first investigated establish suitable ranges that give stable estimate. Using framework with ranges, e-SFR, per ISO 12233, estimated. Initial validation results obtained implementing measuring images linear and non-linear system. For system, closely approximate 12233 measurement. Non-linear measurements exhibit scene-dependant characteristics expected methods. requirements implement systems then discussed.