پردازش تصاویر هوایی: فهرست مسابقات، دیتاست‌ها، مقالات و آموزش‌ها

به شکل کلی یادگیری ماشینی چهار نوع کاربرد در تصاویر هوایی دارد:

  • Image segmentation (cities, roads, water, forest, etc).
  • Object detection (buildings, ships, planes, etc).
  • Resolution enhancement of imagery.
  • Change detection at a site of interest.

منابع ادامه‌ی این نوشته رو سعی می‌کنم توی دسته بندی‌های بالا قرار بدم.

لیست پایان‌نامه‌های انگلیسی:

پیشنهاد پایان‌نامه دانشگاه زوریخ با عنوان «Mowing pattern recognition from Sentinel-​2 images with deep sequence modeling» برای سال ۲۰۱۹

پیشنهاد پایان‌نامه دانشگاه زوریخ با عنوان «Deep sequence modeling for crop classification from Sentinel-​2 satellite images» برای سال ۲۰۱۹

منبع:
https://prs.igp.ethz.ch/education/open-topics-for-students.html

پایان‌نامه کارشناسی ارشد رشته‌ی هوش مصنوعی دانشگاه آمستردام با عنوان «Convolutional Neural Networksfor Crop Yield Predictionusing Satellite Images»

لیست کنفرانس‌های تخصصی مرتبط پردازش تصاویر هوایی:

IN21D: Deep Learning for Geoscience Posters:
https://agu.confex.com/agu/fm18/prelim.cgi/Session/49601

IN14A: Deep Learning for Geoscience I:
https://agu.confex.com/agu/fm18/prelim.cgi/Session/60553

Image segmentation:

مطالب آموزشی:

Satellite images semantic segmentation with deep learning

How to Segment Buildings on Drone Imagery with Fast.ai & Cloud-Native GeoData Tools

مقاله‌ها:


TernausNetV2: Fully Convolutional Network for Instance Segmentation‍
code: https://github.com/ternaus/TernausNetV2

Landscape Classification with Deep Neural Networks
code: https://github.com/dbuscombe-usgs/dl_landscapes_paper

Feedback Neural Network for Weakly SupervisedGeo-Semantic Segmentation

دیتاست‌ها:

DroneDeploy Segmentation Dataset[8GB]

Instance Segmentation:

  • xView 2 Building Damage Asessment Challenge (DIUx, Nov 2019) .
    550k building footprints & 4 damage scale categories, 20 global locations and 7 disaster types (wildfire, landslides, dam collapses, volcanic eruptions, earthquakes/tsunamis, wind, flooding), Worldview-3 imagery (0.3m res.), pre-trained baseline model. Paper: Gupta et al. 2019
  • Microsoft BuildingFootprints Canada & USA (Microsoft, Mar 2019)
    12.6mil (Canada) & 125.2mil (USA) building footprints, GeoJSON format, delineation based on Bing imagery using ResNet34 architecture.
  • Spacenet Challenge Round 4 – Off-nadir (CosmiQ Works, DigitalGlobe, Radiant Solutions, AWS, Dec 2018)
    126k building footprints (Atlanta), 27 WorldView 2 images (0.3m res.) from 7-54 degrees off-nadir angle. Bi-cubicly resampled to same number of pixels in each image to counter courser native resolution with higher off-nadir angles, Paper: Weir et al. 2019
  • Airbus Ship Detection Challenge (Airbus, Nov 2018)
    131k ships, 104k train / 88k test image chips, satellite imagery (1.5m res.), raster mask labels in in run-length encoding format, Kaggle kernels.
  • Open AI Challenge: Tanzania (WeRobotics & Wordlbank, Nov 2018)
    Building footprints & 3 building conditions, RGB UAV imagery – Link to data
  • Netherlands LPIS agricultural field boundaries (Netherlands Department for Economic Affairs)
    294 crop/vegetation catgeories, 780k parcels, yearly dataset for 2009-2018. Open the atom feed downloadlinks with Firefox etc., not Chrome.
  • Denmark LPIS agricultural field boundaries (Denmark Department for Agriculture)
    293 crop/vegetation catgeories, 600k parcels, yearly dataset for 2008-2018
  • CrowdAI Mapping Challenge (Humanity & Inclusion NGO, May 2018)
    Buildings footprints, RGB satellite imagery, COCO data format
  • Spacenet Challenge Round 2 – Buildings (CosmiQ Works, Radiant Solutions, NVIDIA, May 2017)
    685k building footprints, 3/8band Worldview-3 imagery (0.3m res.), 5 cities, SpaceNet Challenge Asset Library
  • Spacenet Challenge Round 1 – Buildings (CosmiQ Works, Radiant Solutions, NVIDIA, Jan 2017)
    Building footprints (Rio de Janeiro), 3/8band Worldview-3 imagery (0.5m res.), SpaceNet Challenge Asset Library

 

Semantic Segmentation:

Object detection:

دیتاست‌ها:

آموزش‌ها:

Detecting Ships in Satellite Imagery

 

Resolution enhancement of imagery:

مطالب آموزشی:

Pushing The Limits of Open Source Data — Enhancing Satellite Imagery through Deep Learning

Change detection at a site of interest:

مطالب آموزشی:

Predicting Landslides using CNNs and Sentinel-2 Data
code: https://github.com/Yichabod/natural_disaster_pred

سایر:

ویدئو آموزشی:

Hands-on Satellite Imagery Analysis | SciPy 2018 Tutorial | Sara Safavi, Dana Bauer

آموزش‌ها:

Image Segmentation: Kaggle experience (Part 1 of 2)

Calculating NDVI for our Area of Interest

مقاله‌:

Tile2Vec: Unsupervised representation learning for spatially distributed data:
https://arxiv.org/pdf/1805.02855.pdf
https://ermongroup.github.io/blog/tile2vec/#mikolov2013efficient

منابع:

http://math.gmu.edu/~rkhatri3/01_Seminar_Introduction.pdf

https://github.com/chrieke/awesome-satellite-imagery-datasets

https://github.com/robmarkcole/satellite-image-deep-learning

You may also like

پاسخی بگذارید

نشانی ایمیل شما منتشر نخواهد شد. بخش‌های موردنیاز علامت‌گذاری شده‌اند *