Skip to main content

capture_images

Skill class for ai.intrinsic.capture_images skill.

Captures sensor images from the specified camera internally using the camera service.

Prerequisites​

This skill does not have any prerequisite.

Usage Example​

To retrieve and consume the capture result contained in the returned capture_data, use the following code:

from intrinsic.perception.public.proto.v1 import capture_result_pb2
from intrinsic.platform.pubsub.python import pubsub

pub_sub = pubsub.PubSub()
kvstore = pub_sub.KeyValueStore(capture_data.capture_result_location.store)
capture_result = capture_result_pb2.CaptureResult()
kvstore.Get(capture_data.capture_result_location.key).Unpack(capture_result)

Parameters​

capture_result_location​

Specifies where to store the capture result. If not provided, defaults to the capture_results in-memory store (this store currently must not be altered by users) and a key corresponding to the hash of the camera identifier. If multiple capture_images skill instances are running with the same camera, by default each skill call will overwrite the others' results. By manually providing a key for each skill instance this can be avoided.

sensor_ids​

Specifies the sensor ids to capture images from. If empty, the skill captures images from all sensors.

post_processing_by_sensor_id​

A map from sensor id to post processing options such as undistortion, cropping, and resizing. If empty, all images will be undistorted by default (if the respective sensor configs contain distortion parameters or the camera distortion is factory-calibrated).

timeout​

Maximum duration to wait for capturing images from the associated camera. If not provided, defaults to 5s.

camera_settings_to_append​

Camera settings to append to the camera settings coming from the camera config for skill execution.

camera_settings_to_wait_for​

Capturing will be repeated until the specified conditional camera settings are met. You may need to increase the timeout, if you expect to wait for a long time. If empty, the skill will execute exactly one capture call.

camera_settings_to_return​

Camera settings to be returned in the result.

log_debug_data​

Whether to log the captured images.

Capabilities​

camera​

Resource with capability CameraConfig

Returns​

capture_data​

The capture data returned by the skill, containing the capture result location and the camera pose. This can be connected to other skills, which can retrieve the capture result using the key value store. See the usage example for more details.

camera_settings​

Camera settings specified by camera_settings_to_return.

Error Code​

The skill does not have error codes yet