This Recommendation provides a parametric object-recognition-ratio-estimation model to check whether an observer in a monitoring center can recognize an object (e.g., a person who jumps out onto the road or debris on the road) while viewing the video taken by an autonomous vehicle’s surveillance camera and delivered to the monitoring center. This object-recognition ratio can be used as an indicator to check whether the surveillance video delivered to the remote monitoring center is of sufficient quality for remote monitoring around the autonomous vehicle. The input used by the model consists of information obtained from video streams and vehicle information. Four different modes, which are called “modes of operation”, can be used for estimating the object-recognition ratio in this Recommendation: – Mode 0: Information obtained from the video stream, such as the video resolution, video bitrate, and video frame rate, in addition to packet loss information. – Mode 1: Information obtained from the video stream and frozen video frame information based on frame inspection. – Mode 2: The same information as mode 0, in addition to vehicle velocity. – Mode 3: The same information as mode 1, in addition to vehicle velocity.