Systems and methods for identifying damage and theft in an aircraft cargo bay
A method for monitoring an aircraft cargo bay using an aircraft interface device may comprise receiving image data of a cargo bay of an aircraft. The method may further comprise detecting movement in the cargo bay, directing a camera to activate in response to that movement, and processing the captured image data. The aircraft interface device may include pre-trained neural network based algorithms configured for event classification. The method may further comprise generating an alarm based on a theft event classification and transmitting the alarm to an aircraft.
1. A method for monitoring an aircraft cargo bay, comprising:
receiving, by an aircraft interface device, image data from a camera in a cargo bay of an aircraft;
detecting, by the aircraft interface device, movement in the cargo bay;
directing, by the aircraft interface device, the camera to activate, wherein the camera is configured to capture the image data of the cargo bay;
processing, by the aircraft interface device, the captured image data for event detection and event classification, including:
performing, by the aircraft interface device, image segmentation and representation;
performing, by the aircraft interface device, image extraction based on localization and segmentation techniques to generate an extracted image from the captured image data;
processing, by the aircraft interface device, the extracted image through a trained deep neural network for an object classification;
evaluating, by the aircraft interface device, a confidence of the object classification; and
receiving, by the aircraft interface device, feedback including additional images, videos, and/or scans in response to the confidence being low;
generating, by the aircraft interface device, an alarm based on a theft event classification;
commanding, by the aircraft interface device, the camera to record video based on the theft event classification; and
marking, by the aircraft interface device, the recording as an event in response to determining that a pre-configured time interval has passed since recording began.
2. The method of claim 1 , wherein the detecting further comprises the aircraft interface device electrically communicating with a motion sensor.
3. The method of claim 2 , wherein the motion sensor is configured to detect movement in the cargo bay.
4. The method of claim 1 , further comprising:
commanding, by the aircraft interface device, the camera to adjust a view of the cargo bay.
5. The method of claim 1 , further comprising:
interfacing, by the aircraft interface device, the aircraft interface device with on ground infrastructure.
6. The method of claim 5 , wherein the interfacing further comprises sending, by the aircraft interface device, the captured image data to the on ground infrastructure, wherein the on ground infrastructure is configured to analyze the captured image data and is further configured to process the captured image data.
7. The method of claim 6 , wherein the interfacing further comprises receiving,
by the aircraft interface device, processed image data from the on ground infrastructure, and synchronizing, by the aircraft interface device, the aircraft interface device with the on ground infrastructure.
8. The method of claim 1 , wherein the processing further comprises removing a background from the captured image data and extracting a scissor cut of an image through semantic segmentation.
9. A method for monitoring an aircraft cargo bay, comprising:
receiving, by an aircraft interface device, image data from a camera in a cargo bay of an aircraft;
detecting, by the aircraft interface device, movement in the cargo bay based on the image data;
directing, by the aircraft interface device, the camera to activate, wherein the camera is configured to capture the image data of the cargo bay;
processing, by the aircraft interface device, the captured image data for event detection and event classification, including:
performing, by the aircraft interface device, image segmentation and representation;
performing, by the aircraft interface device, image extraction based on localization and segmentation techniques to generate an extracted image from the captured image data;
processing, by the aircraft interface device, the extracted image through a trained deep neural network for an object classification;
evaluating, by the aircraft interface device, a confidence of the object classification; and
receiving, by the aircraft interface device, feedback including additional images, videos, and/or scans in response to the confidence being low; and
monitoring, by the aircraft interface device, loading of cargo to prevent damage based on a cargo-loading event classification.
10. The method of claim 9 , wherein the monitoring further comprises monitoring, by the aircraft interface device, movement within the aircraft cargo bay.
11. The method of claim 10 , further comprising:
generating, by the aircraft interface device, a three-dimensional scan of the cargo bay, wherein the generating further comprises generating, by the aircraft interface device, a three-dimensional scan of cargo containers.
12. The method of claim 11 , further comprising:
calculating, by the aircraft interface device, an area between a fuselage wall of the cargo bay and a container based on the three-dimensional scans.
13. The method of claim 12 , further comprising:
comparing, by the aircraft interface device, a calculated area to a minimum separation standard; and
generating, by the aircraft interface device, an alarm based on the calculated area that is less than the minimum separation standard.