This manual explains how to register the extension software "AI Scene Change Detection" (hereinafter referred to as "this product") with the camera, the settings required before starting operation, and how to operate the camera during operation. The camera must be registered before use. Be sure to read the camera's instruction manual when using the camera. The settings shown in this manual may differ depending on the camera model used and the i-PRO Configuration Tool version (hereinafter referred to as "iCT").
This product is software for more convenient use of AI network cameras (sold separately, hereinafter referred to as cameras). The cameras can learn the normal condition and use the alarm notification function when the condition changes.
Alarms can be notify the following external systems:
i-PRO Management Software  (optional)
Network Recorder (optional)
i-PRO Active Guard
Use iCT for the setting of this product.

・For information about iCT, see the following URL. Please use iCT after V3.5 for the setting of this product.

Click here for more information about iCT

・Refer to the following URL for information on the supported models and software versions of this product.

Technical information website

*"<Control number: Cxxxx>" is the number used to search for the information in our support website below.

・This product is required for each camera.

・An SD Memory Card is essential for use with this product. The use of an SD Memory Card for i-PRO devices is recommended.

・If the SD Memory Card is formatted while this product is in use, restart the camera.

This product defines/learns the normal condition and notifies the specified notification destination of the alarm when it is determined to be abnormal. It is useful in several cases, including the following:
To prevent a door for storing important items from being closed or opened.
To quickly grasp the shortage condition on shelves and replenish them.
To search for exhibits and other items when they are taken away from the place where they should be.
To warn about leaving immediately if a car is parked in a parking prohibited area.
This product has the following features.
Learn up to three normal conditions and create a Detector.
Set and learn up to 16 detection areas.
Adjust alarm notification conditions according to the operating environment. See below for details.
4.3 Advanced settings
Backup/upload data stored on an SD Memory Card.
Please refer to the following for the operation visual of this product. In the operation visual, the state in which the door is closed is learned as the normal condition. Light blue indicates normal condition, blue indicates abnormal condition, and red indicates abnormal and alarm condition.
This product is recommended for indoor use. The following is an example of the assumed use environment.(Outdoors, there are large fluctuation factors such as weather, sunshine fluctuations, and seasons, and detection performance may decline.)
The minimum detection size of the product is 5% or more. Adjust the angle og view so that the target is larger than one (1) grid square displayed on the live image.
When determining the angle of view using the System Design Tool, set the guidance parameters from [AI Application]-[General Purpose]-[Select your camera] as follows.
[Target Height] = Height of sensing object from floor + Height of sensing object
[Target Size] = Short side of the object to be detected
[Target Pixels] = Set the lower detection limit to the value below.
360-degree fisheye cameras are not applicable.
[Ex. (2MP(1080p) Camera)]
To detect missing shelves as shown in the figure below, set as follows.

※In this case, set the detection target at the lowest stage of the image to be reflected smaller.

[Target Height] = 0.82 + 0.98 = 1.80ft.
[Target Size] = 0.66ft.
[Target Pixels] = 100pix the lower detection limit
Above is the case where the learning area is the default setting (entire screen). You may be able to detect distant objects by limiting the learning area.
Please refer to the following for the operation visual of this product.
The following cases may cause a missed alarm or false alarm.
The difference in brightness between the background and the subject is small.
Image brightness is low, such as at night.
The subject is out of focus.
The subject is blurred.
The subject is too small or too large.
The subject is white or blackened.
Light conditions, such as outdoors and windows, are easily changed.
External light such as sunlight and car headlights, and reflection of light such as water pools and glass come in.
Strong external light is inserted to create a subject and other shadows.
The camera lens is wet or dirty.
The camera is shaking.
Other precautions are described below.
If you change the camera image settings before or after learning, it may cause a failure or false alarm. When you change the [Image capture mode] or [Image rotation] settings of the camera, perform the following again.
When using in an environment where color images and black-and-white images are switched, it is recommended that you train color images and black-and-white images individually.
Images of illumination assumed during operation must be collected and learned.
When using under IR-LED, you need to train with images under IR-LED.
Enable the [Exclude a person from scene change] setting when you use the camera in environments where people can be seen. However, if a person is not detected correctly, they may not be excluded correctly. If a non-human entity is detected as a human being, it may be excluded.
Enable [Abandonment detection] when using the product for detecting obsolete objects (suspicious objects, etc.).
If a false alarm is generated at low illumination, activating the setting to stop the alarm at low illumination prevents false alarms.
This product cannot be set or trained from multiple PCs at the same time.
If you execute [Reset to the default] of the camera, the setting data of this product and the data in the SD memory card (Learning images, Detector) will not be initialized. However, since it may be affected by the image quality settings of the camera itself, please check the operation again.
The message ID of TCP alarm notification is "59" for AI scene change detection alarm and "60" for the alarm when the SD memory card cannot be accessed.