4.Advanced settings

4.1Detection area setting

4.1.1Detection area setting

Set the detection area.
For details on how to set the detection area, refer to the following.
→3.2　Set the detection area

4.1.2Set the detection area group

Set the detection area group.
Normally, when one of the detection areas meets the detection conditions, the alarm is activated. By setting a group, the alarm is activated when all the detection areas in the group meet the detection conditions.
Set the detection area group as follows.

・

Select [Detection area group number]. You can select a group number from 1 to 4.

・

Set the [Detection area group name] of the detection area group.

・

Select the detection area to include in the group.

・

Click [Set] to apply the setting.

4.1.3Display the demo screen

Check the learning accuracy.

Press [Display] to start the demo screen.
Refer to the following for details about the demo screen.
→5.　Demo screen

4.2Learning

This chapter describes the setup for learning. Refer to the table below for details of the problem and refer to the appropriate chapter.
For information about how to configure [Detector setting] and [Saving learning images], see the following.
→3.　Basic settings


Description	Reference
I want to know the functions of the image list display screen.	→4.2.1　Image selection screen for learning
I want to change the area to learn for each image. I want to detect small objects.	→4.2.2　Learning area settings

4.2.1Image selection screen for learning

This displays the list of images saved for learning and performs learning. Clicking [Display] displays a list of saved images.

・

Select the image you want to learn and click [Start Learning] in the upper left corner to start learning.

・

If you want to delete an image from the list, select the image you want to delete and click [Delete Image] to delete it.

・

An image that has already been learned displays a green check mark at the bottom left of the image.

1.

Click [Start Learning] in the upper left corner to start learning and display the progress of learning in real time.

2.

When selecting a large number of images, you can select and cancel all images by specifying the image selection method from the pull-down on the upper right.
Select from [Select All], [Unselect All], [Select Learned Only], [Select unlearned only], [In-page selection], or [In-page cancellation].

3.

Click the thumbnail image to enlarge the thumbnail image.
Click [Back] in the upper left corner to display the previous image, and [Next] to display the next image.
Click [Delete Image] in the upper right corner to delete the displayed image.

When multiple learning areas are set up, they can be enabled or disabled by clicking in the learning area as shown in the figure below. The disabled state is represented by a dotted line with a white border.

If you do not want to use the area for learning, disable it.

4.

Click the buttons at the bottom to switch pages.
The following operations and displays are explained from left to right.

・

Start page display button: Click to display the first page.

・

Previous Page View button: Click to display the previous page.

・

Page Selection: The selected page is displayed.

・

Last page: The number of the last page is displayed.

・

Next Page View button: Click to display the next page.

・

Last page view button: Click to display the last page.

4.2.2Learning area settings

Set the learning area. Since the set range is used for learning, this is set to change the area to learn for each image or to detect a smaller object.
Set the learning area as follows.

・

Set the learning area on the live screen. By selecting the start and end points, the learning area is set as a rectangle.
By default, all areas are configured and up to four areas can be specified.

・

When set, the unlearned areas appear darker.

・

To delete, select the area you want to delete and click the trash button.
The selected area is displayed with a dotted line.

・

When the magnified thumbnail image is displayed by clicking it on the [Image selection screen for learning], the learning area can be enabled/disabled for each image.
See below for [Image selection screen for learning].
→4.2.1　Image selection screen for learning

The status of the learning area in each image can also be checked on the image list screen.

If the learning area is reset, check whether the learning area is enabled or disabled before re-learning.
Switching the Detector resets the learning area.
For models with [RAM capacity expansion mode] of [Ext. software] on the camera, [RAM capacity expansion mode] must be turned on prior to setting the learning area.

4.3Advanced settings

This chapter describes the settings to be adjusted in case of trouble in the following cases. Please refer to the table below for details of the problem and refer to the appropriate chapter.


Description	Reference
The alarm is not activated due to the intense change in light and darkness.	→4.3.1　Advanced detection settings [Stop alarm in low light conditions]
Areas where people are detected should also be included in the scene change detection.	→4.3.1　Advanced detection settings [Exclude a person from scene change]
If the image is black and white, use a different Detector from that for color images.	→4.3.1　Advanced detection settings [Detector used in black and white video]
Falsely detect or miss the person	→4.3.1　Advanced detection settings [Exclude a person from Scene change]
Adjust the frequency of alarm notification	→4.3.1　Advanced detection settings [Alarm notification operation setting]
Notify HTTP server of alarm	→4.3.2　HTTP notification destination
Notify MQTT server of alarm	→4.3.3　MQTT transmission
Back up/upload data on the SD Memory Card.	→4.3.4　Maintenance screen

4.3.1Advanced detection settings

Make detailed settings related to detection. These settings are common to all detection areas 1 to 16.

1.

[Alarm notification operation setting]

Sets the alarm notification operation when the alarm continues to active.
[Periodic notification]: Alarm notification is performed at [Detection time] interval.
[Continuous notification]: Alarm notification is performed continuously at [Alarm deactivation time] interval of camera setting.
[One-time notification]: Alarm notification is performed only the first time that the alarm condition has been reached.
Default: [Periodic notification]

2.

[Exclude a person from scene change]

Select whether to include the area where the person is detected in the scene change detection.
Default: [On]

When [Exclude a person from scene change] is set to [Off], the person area in the detection area is displayed in red.

When [Exclude a person from scene change] is set to [On], the person area in the detection area is excluded from the scene change.

3.

[Human identification threshold (head)]

Set the detection threshold for the head. The smaller the value, the easier it is to detect as a person, but it is easier to detect incorrectly.
1 to 99
Default: 50

4.

[Human identification threshold (whole body)]

Set the detection threshold for the whole body. The smaller the value, the easier it is to detect as a person, but it is easier to detect incorrectly.
1 to 99
Default: 50

[Human identification threshold (head)] and [Human identification threshold (whole body)] are used for characterization when [Exclude a person from scene change] is [On].

5.

[Detector used in black and white video]

Select the Detector to be used when the input image is black and white. When [None] is selected, the alarm will not be activated.
[None], [1], [2], [3]
Default: None

6.

[Stop alarm in low light conditions]

When set to [On], detection stops when the illuminance is below the set [Low light threshold].
Default: [Off]

7.

[Low light threshold]

Select this if [Stop alarm in low light conditions] is set to [On]. Detection stops when the illumination level falls below the set illumination threshold.
[5 lx]、[10 lx]、[20 lx]、[30 lx]、[40 lx]、[50 lx]
Default: 10 lx

Illumination intensity is estimated from the image, resulting in an error. The estimated illuminance is displayed on the demo screen when [Stop alarm in low light conditions] is turned on. Especially if the brightness differs greatly within the angle of view, the error will be large. For this reason, this function is recommended for use in an environment where the brightness of the entire screen clearly changes, such as when lighting is ON or OFF. Confirm operation in the actual environment before use. When the input image is black and white, detection stops regardless of the illumination intensity.

4.3.2HTTP notification destination

In [HTTP notification destination], set the notification destination (1 to 4) and the notification interval for alarm notification to the HTTP server.

1.

[Transmission destination 1, Transmission destination 2, Transmission destination 3, Transmission destination 4]

Alarm information can be individually notified in up to four locations. The settings of the four notification destinations are the same.

2.

[Transmission]

When the check box is selected, the message is sent.
Default: Unchecked

3.

[Transmission address]

Set the IP address of the notification destination.

4.

[Transmission path name]

Enter the path name of the notification destination. Enter up to 128 half-width alphanumeric characters and symbols.

5.

[SSL]

Select whether to use SSL for notification.
Default: On

6.

[Transmission port number]

Set the port number of the notification destination.
1 to 65535

7.

[Username]

Enter the user name for authentication with the notification destination. Enter up to 128 half-width alphanumeric characters and symbols.

8.

[Password]

Enter the password for authentication with the notification destination. Enter up to 63 half-width alphanumeric characters and symbols.

9.

[Transmission interval]

Select the time interval to notify.
[1 min]、[2 min]、[3 min]、[4 min]、[5 min]

Default: [5 min]

For details of regular HTTP transmissions, refer to the "AI Scene change detection I/F Specifications".
→Technical information and support videos

4.3.3MQTT transmission

[MQTT transmission] configures the destination and send operation to send alarms to the MQTT server.

1.

[Transmission]

When the check box is selected, the message is sent.
Default: Unchecked

2.

[To MQTT transmission setting]

Click this button to display the MQTT setting screen on the camera.

3.

[Topic]

Enter the name of the MQTT topic to be sent. Enter up to 128 half-width alphanumeric characters and symbols.

4.

[QoS]
Select the level of QoS. Signal quality increases as follws: [0]<[1]<[2].

[0]: Messages are delivered up to once in QoS0. Messages are not guaranteed to reach the server.

[1]: QoS1 delivers at least one message. Messages are guaranteed to reach the destination, but may be duplicated.

[2]: QoS2 delivers the message exactly once. Messages are guaranteed to arrive only once without excess or deficiency.

Default: [1]

5.

[Retain]

Select [On] to have the last notified message saved on the MQTT server.
Default: [Off]

For the payload of MQTT, refer to the "AI Scene change detection I/F Specifications".
→Technical information and support videos

4.3.4Maintenance screen

[Maintenance] backs up data stored in the SD Memory Card and initializes the SD Memory Card settings. Clicking [Display] displays [Maintenance].

[Maintenance] downloads/uploads data on SD Memory Cards.

1.

[Download/Upload]
Select the operation to be performed on the SD Memory Card.

-

[Download]: Downloads Detector and Learning images from the SD Memory Card to the PC as zip files. The files to be downloaded are selected in [Download Files].

-

[Upload]: Uploads zipped files generated by [Download] from PC to SD Memory Card. The files to be uploaded are specified by [Upload files].

2.

[Download Files]

Select the data to download from the SD Memory Card. Set when [Download] is selected in [Download/Upload].
[Detector + Image], [Detector only]

3.

[Upload files]

Select the file to be uploaded to the SD Memory Card. The file specified here is the zip file generated by [Download]. Set this when [Upload] is selected in [Download/Upload].

・In [Download], a zip file is generated in the following folder specified on the [Save in] screen of [Option] in iCT.

[Option]-[Save in]\AISceneDetection\yyyyMMdd_hhmmss

・Note that Learning images and Detector stored on the SD Memory Card will be deleted when [Upload] is executed. (If [Detector only] is selected for [Download Files] in [Download], Learning images will not be deleted.)

・For models with [Ext. software] [RAM capacity expansion mode] on the camera, [RAM capacity expansion mode] must be turned on prior to performing [Upload]. (If it cannot be turned ON, follow the manual procedures below.)

・Instead of performing [Upload] in [Maintenance], it is also possible to copy data directly to the SD memory card by following the procedure below.

① Format the SD memory card with the camera

② Insert SD memory card into PC

③ Unzip the file generated by [Download]

④ Create the following folder (the following is when the SD memory card is recognized as the D drive on the PC)

D:\adamapp\

⑤ Copy the folder unzipped in ③ (folder name: AISceneChangeDetection) to the folder created in ④

⑥ Insert the SD memory card into the camera and restart the camera

4.

[Configuration data initialization]

Click [Execute] to initialize all settings.