Difference between motion detection and intrusion detection?

Adkhik

n3wb
Dec 26, 2014
12
0
On a DS-2CD2432F-IW running 5.2, what is the difference between the motion detection and intrusion detection options?
 
Just reading your post and had the same question. Looking in the manual on the disk that came with my camera the differences are below. Still learning myself.

6.6.1 Configuring Motion Detection

Purpose:
Motion detection detects the moving objects in the configured surveillance area, and
triggers the certain action as a respond to detection.
In order to detect the moving objects accurately and reduce the false alarm rate,
normal configuration and expert configuration are selectable for different motion
detection environment.

6.6.8 Configuring Intrusion Detection
Intrusion detection can set an area in the surveillance scene and once the area is been
entered, a set of alarm action is triggered.

Steps:
1. Check the Enable Intrusion Detection checkbox.

2. Click Draw Area, and then draw a rectangle on the image as a defense region.

Note: when you draw the rectangle, all lines should connect end to end to each
other. Up to four areas are supported.
You can click Clear to clear the areas you drawn.
The defense region parameters can be set separately.

3. Choose the region to be configured.
 Threshold: Range [0-10s], the threshold for the time of the object
loitering in the region. If you set the value as 0, alarm is triggered
immediately after the object entering the region.

 Sensitivity: Range [1-100]. The value of the sensitivity defines the size of
the object which can trigger the alarm, when the sensitivity is high, a very
small object can trigger the alarm.

 Percentage: Range [1-100]. Percentage defines the ratio of the in-region
part of the object which can trigger the alarm. For example, when you set
the percentage as 50%, half of the object entering the region will trigger
the alarm.

Arming Schedule is configured to set the time you want the function to be
enabled.

1. Click Edit to set the arming schedule.

2. Choose to trigger alarm actions as Notify Surveillance Center, Send Email,
Upload to FTP and Trigger Channel or trigger the Alarm Output.

3. Click Save to save the settings.
 
The functions still sound similar. Would anyone know the nuances of when you would want to choose intrusion detection over motion detection?
 
Yes, let's say someone approaches the side of your home, no big deal, the camera starts recording, but if they step toward your door or window ( into the threat zone) it can trigger a siren blast or something.
 
This is a key difference:

Threshold: Range [0-10s], the threshold for the time of the object loitering in the region

It would likely minimize false alarms from car headlights at night or heavy snow.
 
I have been using Intrusion Detection the past few days and it has significantly reduced false triggers since you can set the threshold for how long the motion can be present before triggering. The way I have it setup detects all vehicle and human traffic on my driveway and eliminates false triggers caused by car headlights driving past my house and snow/weather changes.
 
  • Like
Reactions: DaveP
Which sensitivity and percentage do you suggest, in order to minimise fals alarms by moving Cat, birds etc?
 
Which sensitivity and percentage do you suggest, in order to minimise fals alarms by moving Cat, birds etc?

The settings are different for each setup because the cat in my camera could closer or further away then the cat would be in yours. You have to work out what they should be for your setup, there is no magic number.
 
Besides that Hikvision user manuals are written in Chenglish which makes it hard to understand which setting does what (at least for me as a non-native English speaker), there are some (main) differences:

  • I think that intrusion detection is only triggered when something crosses into the boundary. Similar as line-crossing (which can be configured uni- or bidirectional). Probably the underlying technique is similar to motion detection, but it only monitors a 'thick border' of the configured boundary, instead of the whole area covered by the boundary.
  • Now, the main difference which makes it unusable for me, is that it triggers just a single event when it detects an intrusion (note that the actual recording is done at my Synology NAS but that doesn’t matter I guess). Say a burglar is crossing the boundary. This is precisely recognized and the recording starts - 5 seconds before and 5 seconds after the alarm as I have set it up. So I get about 15 seconds of footage. Everything what happens next, e.g. the burglar opens a toolbox and drill holes in my locks is not recorded anymore. Nothing is actually recorded anymore unless he steps out of the boundary and in. Motion detection in contrary, keeps recording as long as motion is detected within the area.
  • There are some smaller differences like in basic motion detection you can define an area by small building blocks (draw one block too many means starting over :(), in expert mode you can define up to 8 horizontal aligned rectangles where you can draw only 1 quadrilateral in intrusion mode (why the hell these differences?? :(). I mention ‘only 1’ since the manual mentions ‘up to 4’ while this doesn’t work in my firmware.
 
Last edited by a moderator:
I am not sure that you are correct about the intrusion zone just being a boundary event. The Threshold value in the settings can be set so that the object has to linger within the box for a specific amount of time before the event is recorded, for that to work it also as to consider motion with in the box. I have also had the event set off by wind moving the leaves of a plant within an intrusion box.

What I have noticed is that DS-2CD2332's with v5.2.0 record for a while but then sometimes skip sections of recording but the same type of camera with v5.2.5 does not seem to have the same problem. If I get the time I will upgrade one of my v5.2.0 cameras to see if the problem remains, in which case it could be a hardware variant problem, anyone have a link to a the v5.2.5 firmware that will upgrade a Chinese camera from v5.2.0 and remain multi-lingual without bricking it or having to hex edit?
 
Besides that Hikvision user manuals are written in Chenglish which makes it hard to understand which setting does what (at least for me as a non-native English speaker), there are some (main) differences:

  • I think that intrusion detection is only triggered when something crosses into the boundary. Similar as line-crossing (which can be configured uni- or bidirectional). Probably the underlying technique is similar to motion detection, but it only monitors a 'thick border' of the configured boundary, instead of the whole area covered by the boundary.
  • Now, the main difference which makes it unusable for me, is that it triggers just a single event when it detects an intrusion (note that the actual recording is done at my Synology NAS but that doesn’t matter I guess). Say a burglar is crossing the boundary. This is precisely recognized and the recording starts - 5 seconds before and 5 seconds after the alarm as I have set it up. So I get about 15 seconds of footage. Everything what happens next, e.g. the burglar opens a toolbox and drill holes in my locks is not recorded anymore. Nothing is actually recorded anymore unless he steps out of the boundary and in. Motion detection in contrary, keeps recording as long as motion is detected within the area.
  • There are some smaller differences like in basic motion detection you can define an area by small building blocks (draw one block too many means starting over :(), in expert mode you can define up to 8 horizontal aligned rectangles where you can draw only 1 quadrilateral in intrusion mode (why the hell these differences?? :(). I mention ‘only 1’ since the manual mentions ‘up to 4’ while this doesn’t work in my firmware.

I also don't think it is only when objects cross boundaries. On a windy overcast day I occasionally get intrusion detections from the reflection of the clouds on the roof/wind screen of cars in my driveway.

This may not work for you depending how many cameras you have but I just leave mine on 24/7 record while receiving email notifications for intrusion detections. This way I catch everything and can just skim through the recording if any notifications warrant a closer look.
 
  • Like
Reactions: andakali
One difference I noticed is that only Motion detection includes images of the event in its email alerts - neither line crossing nor intrusion seems to include a photo.
 
Thanks! It looks like doing intrusion detection on the camera includes a photo in the email sent. However, I was doing motion/intrusion detection directly on the Hikvision NVR, which seems to only send a picture if I do motion detection.. I essentially have the same settings set up from my NVR..
 
All of them can send out emails with images attached (talking Hikvison/LTS DVRs). Actually this is a setting in network/email and has nothing to do with events.
 
Yes Hik NVR + Hik cameras. So then why with that setting to include images (in Email settings) do I only get motion detection images by email? With the other two I don't get images, just the emails.
 
....

6.6.8 Configuring Intrusion Detection

....

2. Click Draw Area, and then draw a rectangle on the image as a defense region.

Note: when you draw the rectangle, all lines should connect end to end to each
other. Up to four areas are supported.

....

I know this is an old thread but I can't figure out how to add more than one area for the intrusion detection. The hikvision web interface has a drop down box for Regions but the only option in it is "1". I can see nowhere that lets you add another region so I can draw another area.

AnyJZmx.gif


Anyone have an ideas?