Identifying the Cause of Poor Counts
When checking the counting accuracy, the Event Summary at the bottom of the VT Tuning screen helps you identify the cause of poor counting. Here are possible entries in the Event Summary
| Parameter: | Value: | Meaning: |
| Direction | 0 or 1 | Whether the person is going up or down the screen. 0 is down-screen and 1 is up-screen. |
| Seconds since last Reset | The approximate number of seconds since you clicked the Zone button | |
| Exit Line | 0-7 | If 0 an object is going up-screen and exiting at the top reference line If 7 an object is going down-screen and exiting at the bottom reference line If 1-6 the object has disappeared without crossing the reference line and will not be counted. This may be because of a shadow or a light reflection. Make sure that people passing through have enough contrast throughout the zone. |
| Width | 1-32 | The width of an object in pixels. (1 pixel is 1/32 of the zone width.) Compare this to the MinDetect and MaxDetect valus in the VT settings table and alter these if necessary. |
| Area | The area a person occupies in pixels squared. | |
| Shadow Score | 0-4 | This is a probability score relating to shadow detection, where 0 is definitely not a shadow and 4 is very likely a shadow. If the score is greater than 0, the system tightens detection criteria in its Suggestions list to avoid counting shadows. |
| > MinDetect | 0 or 1 | 1 indicates that the object was greater than the minimum person size and eligible for counting. 0 indicates it wasn't and will not be counted. |
| Transit Time | 0 or 1 | Indicates the time the object took to pass through the zone. 1 = OK, 0 indicates that the object passed through quickly and could possibly be a shadow. |
| All Lines | 0 or 1 | 0 indicates that the reference lines in the zone have been triggered at once by a single object, which can indicate a shadow. Make sure that the zone is large enough that a single large person does not fill it from top to bottom. Alter the zone size or possibly the camera height. |
| Too Narrow | 0 or 1 | 0 indicates that the object in the zone is considered to be too narrow to be a valid count, and is likely to be a shadow. If this is incorrect alter the minimum person size. |
| Sequential Consistency | 0 or 1 | 0 indicates that object's size has changed as it passed through the zone. False readings from this may occur if the camera is placed to count on stairs, where the person gets nearer the camera as they go up the stairs and vice versa. |
| 2*Min Detect | 0 or 1 | 0 indicates that the object is smaller than the system expects an adult to be (ie, an average adult should be more than twice the minimum person size). If the object I smaller than this it may be considered a small child, or other object not to be counted. The width of objects are only measured on the middle line. |
| Shimmer | 0 or 1 | 0 indicates that the object is all dark with no lighter patches, and may be a shadow. 1 indicates that the object has both light and dark areas. |
| Pos | 0 or 1 | 0 indicates that the object is not brighter than the background, which may indicate a shadow. 1 indicates that the object is brighter than the background. |
| End line commit | 0 or 1 | 0 indicates that the object has stopped on the exit reference line. |
| Split | 0 or 1 | The object has split into two parts as it passed through the zone. This may occur on the overlap if the VT unit is monitoring a wide entrance or corridor. |