1 / 5

Using Object Speed in Pixels/Sec

Using Object Speed in Pixels/Sec. This document describes how to: t ranslate true object speed to pixels/second to support: search for objects traveling at some speed alert on objects traveling in some range of speeds t ranslate observed pixels/second speed to true speed to:

kimball
Download Presentation

Using Object Speed in Pixels/Sec

An Image/Link below is provided (as is) to download presentation Download Policy: Content on the Website is provided to you AS IS for your information and personal use and may not be sold / licensed / shared on other websites without getting consent from its author. Content is provided to you AS IS for your information and personal use only. Download presentation by click this link. While downloading, if for some reason you are not able to download a presentation, the publisher may have deleted the file from their server. During download, if you can't get a presentation, the file might be deleted by the publisher.

E N D

Presentation Transcript


  1. Using Object Speed in Pixels/Sec • This document describes how to: • translate true object speed to pixels/second to support: • search for objects traveling at some speed • alert on objects traveling in some range of speeds • translate observed pixels/second speed to true speed to: • determine real world speed of some observed object

  2. Object Speed in SVS • SVS versions below 4.0 determine speed in terms of pixels/second. All alerts and searches use these units. • Note this speed is with respect to the image and will change depending on how far the object is from the camera. • In SVS versions 4.0 and above, cameras can be calibrated to determine the true speed of an object. Uncalibrated cameras must still use pixels/second. • The calibrated speed is with respect to the real world and will be the same no matter how far the object is from the camera Notes • Forensic search uses • V3.6.7 and below: the maximum observed speed of an objectduring any one frame • V4.0 and above: the average speed during the entire object track • Alerts use the average speed during the last several frames before the alert fires • Object speeds as determined by SVS are approximate. Tracking errors, caused by glare, shadows or any number of causes, will cause wide variations in recorded speeds. Single frame (max) speeds can sometimes be unrealistically large. • Speed of objects moving toward or away from the camera will generally be less accurate than objects traveling directly across the camera view.

  3. Process: Translate between pixels/second and real world speed • Mark two points (A & B) in the world along the direction the object will travel • Examples: Front and back of a car, sidewalk cracks, lamp posts • The larger the distance between the points the better • To translate between true speed and apparent (SVS) speed: SVS = S * P / D S = SVS * D / P Where: D = the distance between points A & B in feet(meters) P = the distance in pixels between those points in the image(see next page) S = true object speed in feet(meters)/second SVS = observed object speed in pixels/second as used by SVS

  4. Process: Determine Pixel Distance Between Two Points in Image • Obtain a still image (e.g. jpg) of one video frame • Screen capture or export video frame from VMS • One entire video image, no surrounding frame • Open the image in a video editing tool (e.g. Paint) • Resize the image to 320x240 pixels (WxH) • Different programs have different ways of doing this • Measure the distance between the two points in the image • Some programs show the X/Y coordinates of a point under the mouse pointer • Other programs show the length of a line you can draw between the points Note: images from Microsoft Paint

  5. Shortcut: Alert on fastest/slowest objects • Overview • It is often sufficient to alert on only the very fastest (or slowest) objects in a scene, regardless of their true speed. This method allows you to tune an alert (or search) so that it fires on only the few fastest objects • Method • Configure the alert with your best guess for min speed • Observe the alert for a period of time • Hours to days, depending on traffic level • If the number of alerts is too large, increasethe min speed till the number of alerts is appropriate to the use case. • If there are very few or no alerts, decrease the min speed till some alerts are triggered. • Very few alerts greatly increase the chance the alert will not fire on all the fastest objects Note • This method has the advantage of tuning the number of alerts to the ability of the user to handle them.

More Related