ML Mondays Part 3 raw jpeg data, version 1

Raw images and jpegs for the ML Mondays OBX dataset

obx_labels.zip contains 326 label images in jpeg format. The numeric integers are 63=deep water (optically deep, cannot see the bottom), 128 = shallow water (optically shallow, i.e. can see the bottom) , 193 = aerated water (surf, swash), 255 = dry land

obx_images_1_30.zip
obx_images_31_60.zip
obx_images_61_90.zip
obx_images_91_120.zip
obx_images_121_150.zip
obx_augimages.zip (augmented imagery)

obx_labels.zip (label imagery)

Raw images and jpegs for the ML Mondays OysterNet dataset

oysternet_all_label_images.zip
oysternet_all_images.zip

ML Mondays Part 3 data, version 2

OBX dataset

very high-resolution digital images of coastal environments in the Outer Banks, North Carolina. The imagery and associated label data have been curated especially for this course, as part of the USGS Remote Sensing of Coastal Change Florence Supplemental Project.

This dataset is courtesy of the U.S. Geological Survey. These data were collected as part of the USGS Remote Sensing of Coastal Change Florence Supplemental Project, and processed collectively by Wayne Wight, Jonathan Warrick (USGS-PCSMC), Christopher Sherwood (USGS-WHCMSC), Andrew Ritchie (USGS-PCSMC), Jenna Brown (USGS-MD-DE-DC WSC), Christine Kranenburg (USGS-SPCMSC), Jin-Si Over (USGS-WHCMSC), and Daniel Buscombe (USGS-PCSMC contractor). Although these data have been processed successfully on a computer system at the U.S. Geological Survey (USGS), no warranty expressed or implied is made regarding the display or utility of the data for other purposes, nor on all computer systems, nor shall the act of distribution constitute any such warranty. The USGS or the U.S. Government shall not be held liable for improper or incorrect use of the data described and/or contained herein.

The categories are (and corresponding integer values in the image labels):

• deep water (63)
• white (broke, aerated) water (128)
• shallow water and saturated ground (191)
• dry sandy terrain (255)

data/obx/sample
16 high-resolution images

The following TFRecord format files contain 20 images and multiclass image labels each:
data/obx/obx15-20.tfrec
data/obx/obx14-20.tfrec
data/obx/obx13-20.tfrec
data/obx/obx12-20.tfrec
data/obx/obx11-20.tfrec
data/obx/obx10-20.tfrec
data/obx/obx09-20.tfrec
data/obx/obx08-20.tfrec
data/obx/obx07-20.tfrec
data/obx/obx06-20.tfrec
data/obx/obx05-20.tfrec
data/obx/obx04-20.tfrec
data/obx/obx03-20.tfrec
data/obx/obx02-20.tfrec
data/obx/obx01-20.tfrec
data/obx/obx00-20.tfrec

and this file contains size image labels and image pairs:
data/obx/obx16-6.tfrec

The data are in random order and from different places and times. Image augmentation has been used. The total dataset consists of 146 original images and associated label images (7360 x 4912 pixels), and 180 augmented images and labels, consisting of random rotations (up to 5 degrees), zoom (up to 20%), and width and heights shifts up to 10%

ML Mondays Part 3 data, version 1

Oysternet dataset

https://scholars.duke.edu/display/pub1419444

Aerial UAV color imagery and labels of oyster reefs in shallow water, made publicly available by Duke University researcher Patrick Gray. This dataset, associated with the tool "OysterNet", consists of many small orthomosaics of intertidal oyster reefs and corresponding labels in text format.

OysterNet is described here: https://github.com/patrickcgray/oyster_net

This version of the data consists of a subset of 1000 x 1000 x 3 pixel orthomosaics of oyster reefs and corresponding labels in text (JSON) format. The dataset consists of 820 images, randomly split into 527 training images, 130 validation images, and 163 test images. Each image pixel has a 3-cm spatial resolution, so each scene is 30 x 30 m.

This dataset has only two classes: intertidal oyster reef and background. The labels have been converted from the original json format to binary label image format (where 1=reef, and 0=everything else). A tutorial describing how this was done is available here: https://colab.research.google.com/drive/1icT8sCXxEKuu_chh8uV27mnqs49ZC6VH