Skip to main content

Robust Methods for the Analysis of Images and Videos for Fisheries Stock Assessment Summary of a Workshop (2015) / Chapter Skim
Currently Skimming:

7 Identification and Classification
Pages 50-56

The Chapter Skim interface presents what we've algorithmically identified as the most significant single chunk of text within every page in the chapter.
Select key terms on the right to highlight them within pages of the chapter.


From page 50...
... , David Jacobs, Serge Belongie (Cornell Tech) , and Gunasekaran Seetharaman (Air Force Research Laboratory)
Read the entire page →
From page 51...
... Now, more rapid data acquisition is possible via imaging, but the processing time by humans has increased, resulting in little net improvement in overall analysis time, although the photographic record is very valuable. He described a representative data survey of the Moorea Coral Reef Long-Term Ecological Research Site: 1,250 images were hand annotated with more than 250,000 annotations.
Read the entire page →
From page 52...
... He also described future work, which involves larger-scale use of automatic im age annotation, leveraging CoralNet data, improving recognition techniques, and adding 3D reconstruction. CLASSIFYING LEAVES USING SHAPE David Jacobs, University of Maryland, College Park David Jacobs explained that he uses species identification techniques to identify tree species.
Read the entire page →
From page 53...
... Jacobs is also working to classify birds in a developing application called Birdsnap.3 FINE-GRAINED VISUAL CATEGORIZATION WITH HUMANS IN THE LOOP Serge Belongie, Cornell Tech Serge Belongie described work that is part of a larger effort, known as Visipedia,4 in which users can search a visual encyclopedia by image. Visipedia is designed to be the visual equivalent of Wikipedia.
Read the entire page →
From page 54...
... The Birds-200 data set5 was the first developed by the CalTech Vision Group, and a new data set, CCUB NABirds 700,6 is planned for release later this year and includes input from an ornithologist. Re searchers ask crowdsourced volunteers to label the presence or absence of birds in images, draw bounding boxes around each bird, and label individual parts of the birds (such as breast patterns, crown color, beak shape, and wing color)
Read the entire page →
From page 55...
... TRACKING VEHICLES IN LARGE-SCALE AERIAL VIDEO OF URBAN AREAS Gunasekaran S Seetharaman, Air Force Research Laboratory Gunasekaran Seetharaman explained that the Air Force Research Laboratory (AFRL)
Read the entire page →
From page 56...
... , which is most suitable for low frame-rate imagery, models the target with a set of color, texture, and shape feature descriptors, then computes the match likelihoods for each feature by comparing the target to a local search image through a sliding window. This method helps inject contextual information (i.e., other knowledge about the dynamics of the field of regard)
Read the entire page →

Key Terms

  • david jacobs
  • coral reef
  • david kriegman
  • cornell tech
  • serge belongie

  • kriegman explained
  • data acquisition
  • reef images
  • coral species
  • angel fire

  • northeastern united
  • wide field
  • jacobs noted
  • jacobs explained
  • gunasekaran seetharaman

  • college park
  • onboard computing
  • species identification
  • coral bleaching
  • coral data

  • carbonate reef
  • described afrl
  • people identify
  • identify tree
  • data fusion

  • reef ecosystem
  • identify biological
  • automatic image
  • labeled coral
  • pervasive coral


    • This material may be derived from roughly machine-read images, and so is provided only to facilitate research.
    More information on Chapter Skim is available.