Object Detection
Instance declaration
When you add an Object Detection block to the workspace, the following instance declaration is automatically inserted into the Python code:
object_detection = ObjectDetection(0)
# When there are multiple instances
object_detection_1 = ObjectDetection(1)Parameters
| Name | Kind | Description | Range / Type | Default |
|---|---|---|---|---|
| index | Dropdown option | Instance number (starting from 0) | Integer 0 or greater | 0 |
Select the camera device
Sets the camera to use for object detection.
Parameters
| Name | Kind | Description | Range / Type | Default |
|---|---|---|---|---|
| unit | Dropdown option | Camera device name | System camera label | - |
Python
object_detection = ObjectDetection(0)
object_detection.device('')Load the model
Loads the trained object model. This step is required in order to use the features of the 'Object Detection' module.
Parameters
| Name | Kind | Description | Range / Type | Default |
|---|---|---|---|---|
| wait | Checkbox | Whether to wait for loading to complete | TRUE / FALSE | TRUE |
Python
object_detection = ObjectDetection(0)
object_detection.load_model(wait=True)Set the maximum object count
Sets the maximum number of objects that can be found. The object count ranges from 0 to 10.
Parameters
| Name | Kind | Description | Range / Type | Default |
|---|---|---|---|---|
| data | Input (block) | Maximum object count | Integer 0 to 10 | - |
Python
object_detection = ObjectDetection(0)
object_detection.max_objects(5)Set the confidence threshold
Sets the minimum probability (confidence) for object detection. Only results whose probability (confidence) is at or above this value are shown on screen. The probability (confidence) ranges from 0 to 1.
Parameters
| Name | Kind | Description | Range / Type | Default |
|---|---|---|---|---|
| data | Input (block) | Confidence threshold | Real number 0 to 1 | - |
Python
object_detection = ObjectDetection(0)
object_detection.confidence_threshold(0.5)Detect once
Finds the objects currently on screen and marks them just once.
Parameters
(none)
Python
object_detection = ObjectDetection(0)
object_detection.detect_once()Start / stop continuous detection
Keeps following the objects currently on screen and marks them on the screen.
Parameters
| Name | Kind | Description | Range / Type | Default |
|---|---|---|---|---|
| unit | Dropdown option | Action | start, stop | - |
Python
object_detection = ObjectDetection(0)
# unit = "start"
object_detection.detect_continuous()
# unit = "stop"
object_detection.stop()Show the detection result
Decides whether or not to show the object detection result on the camera screen.
Parameters
| Name | Kind | Description | Range / Type | Default |
|---|---|---|---|---|
| on | Dropdown option | Display ON / OFF | show (on=True), hide (off=False) | TRUE |
Python
object_detection = ObjectDetection(0)
object_detection.display(True)
object_detection.display(False)Object position information
Returns the position/size values of the specified object.
Parameters
| Name | Kind | Description | Range / Type | Default |
|---|---|---|---|---|
| unit | Dropdown option | Object name | 80 object classes ('person', 'bicycle', 'car', 'motorcycle', 'airplane', 'bus', 'cat', 'dog', 'apple', 'cup', 'laptop', 'cell_phone', and so on) | - |
| pos | Dropdown option | Coordinate | x, y | - |
Python
object_detection = ObjectDetection(0)
object_detection.object('person', 'x')
object_detection.object('bicycle', 'y')Object rectangle information
Returns the position/size values of the rectangle around the specified object region.
Parameters
| Name | Kind | Description | Range / Type | Default |
|---|---|---|---|---|
| unit | Dropdown option | Object name | 80 object classes | - |
| pos | Dropdown option | Rectangle information | min_x, max_x, min_y, max_y, width, height, area | - |
Python
object_detection = ObjectDetection(0)
object_detection.object('person', 'width')
object_detection.object('car', 'area')Distance between two objects
Returns the distance between two object classes.
Parameters
| Name | Kind | Description | Range / Type | Default |
|---|---|---|---|---|
| unit1 | Dropdown option | First object name | 80 object classes | - |
| unit2 | Dropdown option | Second object name | 80 object classes | - |
| type | Dropdown option | Kind of distance | distance (omitted or None), horizontal distance (horizontal), vertical distance (vertical) | None |
Python
object_detection = ObjectDetection(0)
object_detection.get_distance('person', 'bicycle') # distance
object_detection.get_distance('person', 'car', 'horizontal') # horizontal distanceObject confidence
The probability (confidence) that the selected object is correct
Parameters
| Name | Kind | Description | Range / Type | Default |
|---|---|---|---|---|
| unit | Dropdown option | Object name | 80 object classes | - |
Python
object_detection = ObjectDetection(0)
object_detection.object_confidence('person')Model state
Returns the loading state of the object model.
It returns 0 if it has not been loaded yet, 1 while it is loading, and 2 once loading is complete.
Parameters
(none)
Python
object_detection = ObjectDetection(0)
object_detection.model_state()Whether an object was detected
Whether an object was found
Parameters
(none)
Python
object_detection = ObjectDetection(0)
object_detection.detected()Was a particular object detected?
Whether the selected object was found
Parameters
| Name | Kind | Description | Range / Type | Default |
|---|---|---|---|---|
| unit | Dropdown option | Object name | 80 object classes | - |
Python
object_detection = ObjectDetection(0)
object_detection.object_detected('person')