Mort
banner
mortcanty.bsky.social
Mort
@mortcanty.bsky.social
Physicist, retired, refugee from X. Current interests: Remote sensing, statistical image analysis, neural networks.
Bomb craters from June 22, 2025 attack on Isfahan nuclear facility. Tomahawk missiles used, no so-called bunker busters. iMAD change detection algorithm and two Sentinel 2 multi-band infrared images (20m) before and after the raid. See github.com/mortcanty/CR... for details on software used.
June 30, 2025 at 8:55 AM
Seaport Activity Monitoring with Sentinel-1 Polarimetric SAR Imagery github.com/mortcanty/CR...
Hoping to track LA/Long Beach port activity with SAR change detection into the Trumpian Tariff Era. Longer term project most likely/unfortunately.
May 6, 2025 at 3:06 PM
Localizing the extent of the April 24 explosion at a Russian ammunition depot near Moscow. Left: SAR-1 reflectance changes between 5 and 17 April at the 0.1% significance level. Right: significant changes between 17 and 29 April. Methodology: github.com/mortcanty/CRC5Docker
April 30, 2025 at 10:48 AM
Urban sprawl as seen with building object recognition on a U-Net fully convolutional network. Left: from a National Agriculture Image Program (NAIP) image from 2012 over an area west of Houston, Texas. Right: from a NAIP image from 2022 but keeping the 2012 NAIP image as background.
March 11, 2025 at 7:38 PM
Just received the final proofs for the 5th Ed. of my monograph on Image Analysis (CRC Press T&F, 2025). The accompanying software (Jupyter notebooks) is already on github.com/mortcanty/CR... and can be run from a Docker container. You can even query the contents with an LLM/RAG if you like.
GitHub - mortcanty/CRC5Docker: Python scripts and Jupyter Notebooks for the textbook "Image Analysis, Classification and Change Detection in Remote Sensing, Fifth Revised Edition"
Python scripts and Jupyter Notebooks for the textbook "Image Analysis, Classification and Change Detection in Remote Sensing, Fifth Revised Edition" - mortcanty/CRC5Docker
github.com
February 17, 2025 at 6:58 PM
The 5th ed. of my book on image analysis will come out soon. Updated software is already available at github.com/mortcanty/CR...
Just for fun, and because it's "in", I added a LLM/RAG notebook to query the textbook's content informally. Here's poor old Deepseek trying to understand what I want:
February 4, 2025 at 1:53 PM
According to an S1 change detection time series, the destruction of the SOS Hermann Gmeiner School, Rafah (left hand image) occurred between August 7 and 19, 2024 (central image). The AOI used is shown on the right.
January 18, 2025 at 3:49 PM
Palisades fire. Had to wait for latest GEE S1 image. Left: changes over Dec 16-Dec 28. Right: Dec 28-Jan 9. Cyan signifies significant reflectance decreases in VV, VH bands, red increases in VV, VH, yellow increase in one decrease in other. Useful for locating fully destroyed houses? Ground truth!
January 11, 2025 at 12:49 PM
The left image appears* to pinpoint quite severe building damage in central Gaza City in the second half of January, 2024 using Sentinel-1 change detection. The right image plots the relative change frequencies from August, 2023 to now over the same area.

*Being careful, no ground truth.
January 4, 2025 at 4:46 PM
Another example of semantic segmentation with a UNet FCN, this time of a mixture of large and small buildings (part of the Lawrence Livermore Laboratory). See
github.com/mortcanty/CR...
and /plug the forthcoming revised ed. of my textbook on image analysis. /endplug.
December 29, 2024 at 12:18 PM
The color-coded frequency (or total number) of changes at the port of Odessa, as detected with S1 A and S1 B SAR for the period Jan, 2021 to Aug, 2022 (80 images). Water surfaces are always noisy due to wave effects. Therefore the minimum frequency displayed was set to 5 significant changes.
December 27, 2024 at 2:47 PM
Ukraine, June 2023, the Kakhovka Dam is breached causing draining of the Zaporizhzhia nuclear power plant's main water source, the Kakhovka Reservoir, about 150 km upstream. This S1 change image shows in red the reduction in VV, VH reflectance. The Dnieper is still connected to the cooling pond.
December 23, 2024 at 4:49 PM
The Malibu Franklin Fire Disaster seen with the iMAD change detection algorithm on two Sentinel-2 images (Nov 8 and Dec 18, 2024.) RGB color composites: left using the 20m NIR/SWIR bands as input to iMAD, right using the 10m VNIR bands.
github.com/mortcanty/CR...
December 20, 2024 at 6:54 PM
Statistically significant changes in a time series of 27 Sentinel-1A images over Cologne-Bonn Airport, Jan-Dec, 2024. Blue pixels: about 5 changes, red pixels: at least 16 changes detected. Change frequencies below 5 masked out. To and fro at boarding gates account for most of the observed changes.
December 18, 2024 at 3:11 PM
Sentinel-1C was just launched so let's hope that high frequency, weather independent observations can continue. Here for fun is another time series of changes in Sentinel-1
radar reflectance, this time from a high school parking area in Houston, Texas from Nov, 2017 through July, 2020.
December 6, 2024 at 9:51 AM
Sentinel-1 sequential polarimetric change detection of suburban development (Houston, 2022). Red points in lower change frequency plot are reflectance increases, blue points decreases. The initial blue peak is characteristic of vegetation clearing prior to construction.
December 3, 2024 at 3:16 PM
Siege of Mariupol lasted from 2.2.22 til 20.5.22.
1st pic shows significant changes detected in Azovstal Works with polarized Sentinel-1 radar from 1.5.21 to 31.12.22.
2nd plots fraction of changes for approx 6d time intervals. Red: increased VV and VH reflectance. Blue: decreased.
shorturl.at/W6UPv
December 3, 2024 at 10:45 AM
Clear cutting (red pixels) of coniferous forest between June 2021 and June 2022 in Northrhine Westphalia, Landkreis Olpe, necessitated by severe drought conditions. Analysis method: Change detection with the iMAD algorithm followed by k-means clustering, see
developers.google.com/earth-engine...
November 29, 2024 at 2:03 PM
Here is an example of semantic segmentation of man-made structures (suburban housing in this case) with a FCN using UNET architecture. The houses missed were built after the GEE image was archived. See my shared COLAB notebook at

colab.research.google.com/drive/136YAV...

for the details.
November 23, 2024 at 1:35 PM
And to two tutorials on change detection with optical and radar satellites:
developers.google.com/earth-engine...

developers.google.com/earth-engine...
Change Detection in Google Earth Engine - The MAD Transformation (Part 1)  |  Google for Developers
developers.google.com
November 12, 2024 at 1:09 PM
And here's a link to my blog on remote sensing and statistics: fwenvi-idl.blogspot.com
Fun with Python
fwenvi-idl.blogspot.com
November 12, 2024 at 1:07 PM
Just submitted the MS for the fifth revised edition of my textbook "Image Analysis, Classification and Change Detection in Remote Sensing, With Algorithms in Python", CRC Press, Taylor and Francis, 2025. Fourth Edition appeared 2019. See github.com/mortcanty/CR...
GitHub - mortcanty/CRC5Docker: Python scripts and Jupyter Notebooks for the textbook "Image Analysis, Classification and Change Detection in Remote Sensing, Fifth Revised Edition"
Python scripts and Jupyter Notebooks for the textbook "Image Analysis, Classification and Change Detection in Remote Sensing, Fifth Revised Edition" - mortcanty/CRC5Docker
github.com
November 12, 2024 at 12:13 PM