Kaeli Jeffrey W.

No Thumbnail Available
Last Name
Kaeli
First Name
Jeffrey W.
ORCID

Filters

2
1 1
2013 - 2017 2
1 1
2
Reset filters

Settings

Search Results

Now showing 1 - 2 of 2
  • Thesis
    Computational strategies for understanding underwater optical image datasets
    (Massachusetts Institute of Technology and Woods Hole Oceanographic Institution, 2013-09) Kaeli, Jeffrey W.
    A fundamental problem in autonomous underwater robotics is the high latency between the capture of image data and the time at which operators are able to gain a visual understanding of the survey environment. Typical missions can generate imagery at rates hundreds of times greater than highly compressed images can be transmitted acoustically, delaying that understanding until after the vehicle has been recovered and the data analyzed. While automated classification algorithms can lessen the burden on human annotators after a mission, most are too computationally expensive or lack the robustness to run in situ on a vehicle. Fast algorithms designed for mission-time performance could lessen the latency of understanding by producing low-bandwidth semantic maps of the survey area that can then be telemetered back to operators during a mission. This thesis presents a lightweight framework for processing imagery in real time aboard a robotic vehicle. We begin with a review of pre-processing techniques for correcting illumination and attenuation artifacts in underwater images, presenting our own approach based on multi-sensor fusion and a strong physical model. Next, we construct a novel image pyramid structure that can reduce the complexity necessary to compute features across multiple scales by an order of magnitude and recommend features which are fast to compute and invariant to underwater artifacts. Finally, we implement our framework on real underwater datasets and demonstrate how it can be used to select summary images for the purpose of creating low-bandwidth semantic maps capable of being transmitted acoustically.
  • Article
    Do AUVs dream of electric eels?
    (Oceanography Society, 2017-06) Kaeli, Jeffrey W.
    Free-swimming autonomous underwater vehicles (AUVs) are distinct from tethered remotely operated vehicles (ROVs) and human-occupied vehicles in the amount of data-driven feedback a human can provide during a mission. While free-space optical communications afford tether-equivalent data rates at relatively close ranges (Farr et al., 2010), most AUVs employ acoustic modems to maintain two-way communications with their operators while underway (Freitag et al., 2005). However, the low bandwidths and high latencies inherent in underwater acoustics prohibit the real-time transmission of data generated by imaging sensors such as cameras and side-scan sonars. This has profound implications with regard to the meaning of the data an AUV collects and the trust an operator has in the AUV’s autonomy to react to data in the absence of direct human oversight.