|
Guest >> Sign In |
|
|
DIRECTED
ENERGY
PROFESSIONAL
SOCIETY
Abstract: 25-Symp-127
|
|
|
|
UNCLASSIFIED, PUBLIC RELEASE
Atmospheric Sensing Autonomy by Single-Shot Object Detection
Sensor autonomy is crucial in long-term system planning. Passive atmospheric sensors, like DELTA-Wx, rely on distinct features to perform their functions, making them sensitive to sensor drift and adaptive scenes. Traditional computer vision techniques can satisfy some level of sensor autonomy but often struggle in cluttered and complex environments. To meet these system dependencies regardless of environment, we implement a Convolutional Neural Network (CNN) based object detection model to support real-time Modulation Transfer Function (MTF) measurements. Traditional template matching performance is compared to the CNN performance.
UNCLASSIFIED, PUBLIC RELEASE
|