Open Hub
Open Hub

 Settings |  Report Duplicate

0
I Use This!
Very Low Activity

Commits : Listings

Analyzed about 3 hours ago. based on code collected 1 day ago.
Nov 05, 2025 — Nov 05, 2026
Commit Message Contributor Files Modified Lines Added Lines Removed Code Location Date
Updated section 1422
Srinivasan Kannan
More... over 3 years ago
1425. (THEORY and FEATURE) GIS Urban Sprawl Analytics - Four colored morphological settlement classification from R2023A GHS-BUILT-C dataset - Expanded Chennai Metropolitan Area Bounding Box - 12 May 2023 - related to 823,1163,1164,1174,1175,1177,1179,1194,1211,1226,1234,1290,1291,1292,1295,1296,1297,1301,1302,1304,1312,1313,1320,1323,1326,1328,1342,1343,1349,1350,1351,1362,1363,1377,1386,1395,1397,1398,1412,1414,1423,1424 and all sections on GIS Urban Sprawl Analytics, Merit of Large Scale Visuals, Gridded Population Estimators (SEDAC GPW4 and GHSL convex hull estimators), DBSCAN and Neural Network Urban Extent Learners, Computational Geometry, Intrinsic Merit Rankings, Verlhulste-R Ricker Population Estimation, Community detection, Social Network Analytics, Space filling by Curves, Unequal Circle and Polynomial packing, Computational Geometry and Polygonal packing by Voronoi Tessellation, Four color theorem, Economic Merit, Computational Economics - GDP estimation from Urban sprawl demographics, electricity consumption, Poverty alleviation, Preferential Attachment, Automatic Delineation of Urban Sprawls, Cellular Automata-SLEUTH-Polya Urn-Dynamic Facegraph Urban Growth Models,Expander Graphs,Timeseries analysis,Graph Edit Distance and Earth mover distance, Face recognition,Graph Isomorphism (Exact, Approximate and Subgraph),Graph matching,Graph similarity,Image Search,Quadrilateral Mesh Deformation and Isomorphism,Maximum Common Subgraph Problem and Approximate Topological Matching
Srinivasan Kannan
More... over 3 years ago
------------------------------------------------------------------------------------------------------------------------------------------- 1425. (THEORY and FEATURE) GIS Urban Sprawl Analytics - Four colored morphological settlement classification from R2023A GHS-BUILT-C dataset - Expanded Chennai Metropolitan Area Bounding Box - 12 May 2023 - related to 823,1163,1164,1174,1175,1177,1179,1194,1211,1226,1234,1290,1291,1292,1295,1296,1297,1301,1302,1304,1312,1313,1320,1323,1326,1328,1342,1343,1349,1350,1351,1362,1363,1377,1386,1395,1397,1398,1412,1414,1423,1424 and all sections on GIS Urban Sprawl Analytics, Merit of Large Scale Visuals, Gridded Population Estimators (SEDAC GPW4 and GHSL convex hull estimators), DBSCAN and Neural Network Urban Extent Learners, Computational Geometry, Intrinsic Merit Rankings, Verlhulste-R Ricker Population Estimation, Community detection, Social Network Analytics, Space filling by Curves, Unequal Circle and Polynomial packing, Computational Geometry and Polygonal packing by Voronoi Tessellation, Four color theorem, Economic Merit, Computational Economics - GDP estimation from Urban sprawl demographics, electricity consumption, Poverty alleviation, Preferential Attachment, Automatic Delineation of Urban Sprawls, Cellular Automata-SLEUTH-Polya Urn-Dynamic Facegraph Urban Growth Models,Expander Graphs,Timeseries analysis,Graph Edit Distance and Earth mover distance, Face recognition,Graph Isomorphism (Exact, Approximate and Subgraph),Graph matching,Graph similarity,Image Search,Quadrilateral Mesh Deformation and Isomorphism,Maximum Common Subgraph Problem and Approximate Topological Matching ------------------------------------------------------------------------------------------------------------------------------------------- 1425.1 python-src/image_pattern_mining/ImageNet/GISUrbanSprawlAnalytics.py has been updated to include a dictionary of longitude-latitude to data of the raster at that long-lat and return the dictionary (longlatstat) within the return tuple value. This crucial functionality was lacking thus far and enables lookup of raster at longitude-latitude granularity. 1425.2 New function four_colored_morphological_settelement_zones(heightrasterdata,longx,latx,longy,laty) has been defined which iterates through building height raster data (GHSL BUILT-C MSZ dataset is implied unless otherwise stated) and categorizes the pixels by their morphological settlement zone (MSZ) which are broadly classified into vegetation-water,residential and non-residential segments (and their subsegments by height and NDVI). By this four colored urban sprawl facegraph theory is minimally implemented in NeuronRain if non-residential segment is split into industrial and commercial zones by building heights (a filter based on height standards might have to be added - Industrial buildings (manufacturing) have a standard clearing height of 32-36 feet - https://methodarchitecture.com/industrial-clear-height-36/). Function four_colored_morphological_settlement_zones() is invoked on testlogs/RemoteSensingGIS/GHS_BUILT_C_MSZ_E2018_GLOBE_R2023A_54009_10_V1_0_R8_C26.tif GHSL BUILT-C R2023A morphological settlement zones dataset (downloadable from https://jeodpp.jrc.ec.europa.eu/ftp/jrc-opendata/GHSL/GHS_BUILT_C_GLOBE_R2023A/GHS_BUILT_C_MSZ_E2018_GLOBE_R2023A_54009_10/V1-0/tiles/ - 183 MB - not part of the commit): rheightdata=urban_sprawl_from_raster(79.271851,12.439259,80.351257,13.568572,"testlogs/RemoteSensingGIS/GHS_BUILT_C_MSZ_E2018_GLOBE_R2023A_54009_10_V1_0_R8_C26.tif",dt="BUILT_C R2023A") four_colored_morphological_settelement_zones(rheightdata,79.271851,12.439259,80.351257,13.568572) 1425.3 Following are the four colored morphological settlement zone statistics computed by four_colored_morphological_settelement_zones() for earlier Chennai metropolitan area bounding box (http://bboxfinder.com/#12.439259,79.271851,13.568572,80.351257) captured in testlogs/GISUrbanSprawlAnalytics.log.12May2023: urbansprawl_gini_coefficient(): Gini Index of the dataset = 0.9399500643202778 Morphological settlement zones : defaultdict(<class 'int'>, {0: 11017, 11: 265, 2: 229, 12: 58, 3: 223, 5: 43, 13: 213, 1: 105, 23: 8, 22: 1, 14: 36, 24: 2, 4: 2, 25: 1, 15: 1}) 11 : MSZ, built spaces, residential, building height <= 3m: 265 02 : MSZ, open spaces, medium vegetation surfaces 0.3 < NDVI <=0.5: 229 12 : MSZ, built spaces, residential, 3m < building height <= 6m: 58 03 : MSZ, open spaces, high vegetation surfaces NDVI > 0.5: 223 05 : MSZ, open spaces, road surfaces: 43 13 : MSZ, built spaces, residential, 6m < building height <= 15m: 213 01 : MSZ, open spaces, low vegetation surfaces NDVI <= 0.3: 105 23 : MSZ, built spaces, non-residential(commercial,IT-ITES-Industrial), 6m < building height <= 15m: 8 22 : MSZ, built spaces, non-residential(commercial,IT-ITES-Industrial), 3m < building height <= 6m: 1 14 : MSZ, built spaces, residential, 15m < building height <= 30m: 36 24 : MSZ, built spaces, non-residential(commercial,IT-ITES-Industrial), 15m < building height <= 30m: 2 04 : MSZ, open spaces, water surfaces LAND < 0.5: 2 25 : MSZ, built spaces, non-residential(commercial,IT-ITES-Industrial), building height > 30m: 1 15 : MSZ, built spaces, residential, building height > 30m: 1
Srinivasan Kannan
More... over 3 years ago
Section 1411 updated
Srinivasan Kannan
More... over 3 years ago
+1424. (THEORY and FEATURE) GIS Urban Sprawl Analytics - Global Human Settlement Layer 3D Urban Growth Model from R2023A GHS-BUILT-V and GHS-BUILT-S datasets - Expanded Chennai Metropolitan Area Bounding Box - 11 May 2023 - related to 823,1163,1164,1174,1175,1177,1179,1194,1211,1226,1234,1290,1291,1292,1295,1296,1297,1301,1302,1304,1312,1313,1320,1323,1326,1328,1342,1343,1349,1350,1351,1362,1363,1377,1386,1395,1397,1398,1412,1414,1423 and all sections on GIS Urban Sprawl Analytics, Merit of Large Scale Visuals, Gridded Population Estimators (SEDAC GPW4 and GHSL convex hull estimators), DBSCAN and Neural Network Urban Extent Learners, Computational Geometry, Intrinsic Merit Rankings, Verlhulste-R Ricker Population Estimation, Community detection, Social Network Analytics, Space filling by Curves, Unequal Circle and Polynomial packing, Computational Geometry and Polygonal packing by Voronoi Tessellation, Four color theorem, Economic Merit, Computational Economics - GDP estimation from Urban sprawl demographics, electricity consumption, Poverty alleviation, Preferential Attachment, Automatic Delineation of Urban Sprawls, Cellular Automata-SLEUTH-Polya Urn-Dynamic Facegraph Urban Growth Models,Expander Graphs,Timeseries analysis,Graph Edit Distance and Earth mover distance, Face recognition,Graph Isomorphism (Exact, Approximate and Subgraph),Graph matching,Graph similarity,Image Search,Quadrilateral Mesh Deformation and Isomorphism,Maximum Common Subgraph Problem and Approximate Topological Matching ^M
Srinivasan Kannan
More... over 3 years ago
------------------------------------------------------------------------------------------------------------------------------------------- 1424. (THEORY and FEATURE) GIS Urban Sprawl Analytics - Global Human Settlement Layer 3D Urban Growth Model from R2023A GHS-BUILT-V and GHS-BUILT-S datasets - Expanded Chennai Metropolitan Area Bounding Box - 11 May 2023 - related to 823,1163,1164,1174,1175,1177,1179,1194,1211,1226,1234,1290,1291,1292,1295,1296,1297,1301,1302,1304,1312,1313,1320,1323,1326,1328,1342,1343,1349,1350,1351,1362,1363,1377,1386,1395,1397,1398,1412,1414,1423 and all sections on GIS Urban Sprawl Analytics, Merit of Large Scale Visuals, Gridded Population Estimators (SEDAC GPW4 and GHSL convex hull estimators), DBSCAN and Neural Network Urban Extent Learners, Computational Geometry, Intrinsic Merit Rankings, Verlhulste-R Ricker Population Estimation, Community detection, Social Network Analytics, Space filling by Curves, Unequal Circle and Polynomial packing, Computational Geometry and Polygonal packing by Voronoi Tessellation, Four color theorem, Economic Merit, Computational Economics - GDP estimation from Urban sprawl demographics, electricity consumption, Poverty alleviation, Preferential Attachment, Automatic Delineation of Urban Sprawls, Cellular Automata-SLEUTH-Polya Urn-Dynamic Facegraph Urban Growth Models,Expander Graphs,Timeseries analysis,Graph Edit Distance and Earth mover distance, Face recognition,Graph Isomorphism (Exact, Approximate and Subgraph),Graph matching,Graph similarity,Image Search,Quadrilateral Mesh Deformation and Isomorphism,Maximum Common Subgraph Problem and Approximate Topological Matching ------------------------------------------------------------------------------------------------------------------------------------------- 1424.1 python-src/image_pattern_mining/ImageNet/GISUrbanSprawlAnalytics.py has been updated to change plot_surface() in function three_dimensional_urban_growth_model() to solid surface instead of colormap. surf = ax.plot_surface(X, Y, heights) 1424.2 3D Urban Growth Model (UGM) from GHSL R2023A BUILT-V (3D Volume) and BUILT-S (2D surface) have been computed which refine earlier R2022A 3D UGM: r2ddata=urban_sprawl_from_raster(79.271851,12.439259,80.351257,13.568572,"testlogs/RemoteSensingGIS/GHS_BUILT_S_E2030_GLOBE_R2023A_54009_1000_V1_0_R8_C26.tif",dt="BUILT_S R2023A") r3ddata=urban_sprawl_from_raster(79.271851,12.439259,80.351257,13.568572,"testlogs/RemoteSensingGIS/GHS_BUILT_V_E2030_GLOBE_R2023A_54009_1000_V1_0_R8_C26.tif",dt="BUILT_V R2023A") three_dimensional_urban_growth_model(r2ddata,r3ddata,79.271851,12.439259,80.351257,13.568572) 1424.3 Following are maximum-mean-median-standard deviation of building heights computed as per GHSL R2023A definitions in sections 2.2.1 and 2.3 of https://ghsl.jrc.ec.europa.eu/documents/GHSL_Data_Package_2023.pdf?t=1683540422 ["...... The built-up volume (BUVOL) is the volume of the built space above ground, expressed in cubic meters. Be Sx, the surface of the grid sample x. The Average of the Net Building Height of the sample x (ANBHx) is defined as: ANBHx = BUVOLx/BUSURFx (or) BUVOLx = BUSURFx ∗ ANBHx .The Average of the Gross Building Height of the sample x (AGBHx) is defined as: AGBHx = BUVOLx/Sx, BUVOLx = Sx ∗ AGBHx. Developing the above, the BUFRACx can be derived as AGBHx/ANBHx = BUSURFx/Sx = BUFRACx. ..... The data is made by application of the formula: BUVOLx = ANBHx ∗ BUSURFx, where: the ANBHx is the GHS-BUILT-H R2023A - GHS building height, derived from AW3D30, SRTM30, and Sentinel- 2 composite (2018), the BUSURFx is the GHS-BUILT-S R2023A - GHS built-up surface spatial raster dataset, derived from Sentinel-2 composite and Landsat, multitemporal (1975-2030). ....."] : three_dimensional_urban_growth_model(): maximum building height = 15.826443260823885 three_dimensional_urban_growth_model(): mean(heights) = 2.472475289275445 three_dimensional_urban_growth_model(): median(heights) = 2.50044877498693 three_dimensional_urban_growth_model(): stdev(heights) = 1.9540860578186092 1424.4 Changed 3D UGM Digital Elevation Model graphics (bird's eye view) for R2023A BUILT-V Chennai Metropolitan Area Bounding Box has been updated at testlogs/GISUrbanSprawlAnalytics.ChennaiMetropolitanArea_R2023A_3D_DEM.11May2023.jpeg (rotated in matplotlib3d viewer and snapshotted to JPEG) 1424.5 BUFRAC or Floor-Space Index could be arrived at on similar lines if surface area Sx is accurately known. GHSL R2023A BUILT-C datasets in https://jeodpp.jrc.ec.europa.eu/ftp/jrc-opendata/GHSL/GHS_BUILT_C_GLOBE_R2023A/GHS_BUILT_C_MSZ_E2018_GLOBE_R2023A_54009_10/V1-0/tiles/GHS_BUILT_C_MSZ_E2018_GLOBE_R2023A_54009_10_V1_0_R8_C26.zip directly classify building heights by morphological settlement zones (which cover 3 of the 4-color classification of urban sprawl zones - vegetation-water,residential and non-residential) as per following legend from https://ghsl.jrc.ec.europa.eu/ghs_buC2023.php but due to huge size of the datasets (12 GB in total) GHSL BUILT-C analytics are withheld (atleast till there is an algorithm to segregate non-residential spaces into industrial and commercial segments by building heights or any other GIS data, thereby making it a 4-coloring): 01 : MSZ, open spaces, low vegetation surfaces NDVI <= 0.3 02 : MSZ, open spaces, medium vegetation surfaces 0.3 < NDVI <=0.5 03 : MSZ, open spaces, high vegetation surfaces NDVI > 0.5 04 : MSZ, open spaces, water surfaces LAND < 0.5 05 : MSZ, open spaces, road surfaces 11 : MSZ, built spaces, residential, building height <= 3m 12 : MSZ, built spaces, residential, 3m < building height <= 6m 13 : MSZ, built spaces, residential, 6m < building height <= 15m 14 : MSZ, built spaces, residential, 15m < building height <= 30m 15 : MSZ, built spaces, residential, building height > 30m 21 : MSZ, built spaces, non-residential, building height <= 3m 22 : MSZ, built spaces, non-residential, 3m < building height <= 6m 23 : MSZ, built spaces, non-residential, 6m < building height <= 15m 24 : MSZ, built spaces, non-residential, 15m < building height <= 30m 25 : MSZ, built spaces, non-residential, building height > 30m
Srinivasan Kannan
More... over 3 years ago
Section 1423 updated
root
More... over 3 years ago
GHSL R2023A SMOD Visualization from https://ghsl.jrc.ec.europa.eu/visualisation.php#[email protected],79.84039,9z&v=301&ln=1&gr=ds&lv=10000000000000000000000000000000000000011111&lo=aaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaa&pg=V
root
More... over 3 years ago
+1423. (THEORY and FEATURE) GIS Urban Sprawl Analytics - Global Human Settlement Layer Degree of Urbanization comparison between R2022A and R2023A GHS-SMOD datasets - Expanded Chennai Metropolitan Area Bounding Box - 9 May 2023 - related to 823,1163,1164,1174,1175,1177,1179,1194,1211,1226,1234,1290,1291,1292,1295,1296,1297,1301,1302,1304,1312,1313,1320,1323,1326,1328,1342,1343,1349,1350,1351,1362,1363,1377,1386,1395,1397,1398,1412,1414 and all sections on GIS Urban Sprawl Analytics, Merit of Large Scale Visuals, Gridded Population Estimators (SEDAC GPW4 and GHSL convex hull estimators), DBSCAN and Neural Network Urban Extent Learners, Computational Geometry, Intrinsic Merit Rankings, Verlhulste-R Ricker Population Estimation, Community detection, Social Network Analytics, Space filling by Curves, Unequal Circle and Polynomial packing, Computational Geometry and Polygonal packing by Voronoi Tessellation, Four color theorem, Economic Merit, Computational Economics - GDP estimation from Urban sprawl demographics, electricity consumption, Poverty alleviation, Preferential Attachment, Automatic Delineation of Urban Sprawls, Cellular Automata-SLEUTH-Polya Urn-Dynamic Facegraph Urban Growth Models,Expander Graphs,Timeseries analysis,Graph Edit Distance and Earth mover distance, Face recognition,Graph Isomorphism (Exact, Approximate and Subgraph),Graph matching,Graph similarity,Image Search,Quadrilateral Mesh Deformation and Isomorphism,Maximum Common Subgraph Problem and Approximate Topological Matching ^M
Srinivasan Kannan
More... over 3 years ago
-------------------------------------------------------------------------------------------------------------------------------------------- 1423. (THEORY and FEATURE) GIS Urban Sprawl Analytics - Global Human Settlement Layer Degree of Urbanization comparison between R2022A and R2023A GHS-SMOD datasets - Expanded Chennai Metropolitan Area Bounding Box - 9 May 2023 - related to 823,1163,1164,1174,1175,1177,1179,1194,1211,1226,1234,1290,1291,1292,1295,1296,1297,1301,1302,1304,1312,1313,1320,1323,1326,1328,1342,1343,1349,1350,1351,1362,1363,1377,1386,1395,1397,1398,1412,1414 and all sections on GIS Urban Sprawl Analytics, Merit of Large Scale Visuals, Gridded Population Estimators (SEDAC GPW4 and GHSL convex hull estimators), DBSCAN and Neural Network Urban Extent Learners, Computational Geometry, Intrinsic Merit Rankings, Verlhulste-R Ricker Population Estimation, Community detection, Social Network Analytics, Space filling by Curves, Unequal Circle and Polynomial packing, Computational Geometry and Polygonal packing by Voronoi Tessellation, Four color theorem, Economic Merit, Computational Economics - GDP estimation from Urban sprawl demographics, electricity consumption, Poverty alleviation, Preferential Attachment, Automatic Delineation of Urban Sprawls, Cellular Automata-SLEUTH-Polya Urn-Dynamic Facegraph Urban Growth Models,Expander Graphs,Timeseries analysis,Graph Edit Distance and Earth mover distance, Face recognition,Graph Isomorphism (Exact, Approximate and Subgraph),Graph matching,Graph similarity,Image Search,Quadrilateral Mesh Deformation and Isomorphism,Maximum Common Subgraph Problem and Approximate Topological Matching -------------------------------------------------------------------------------------------------------------------------------------------- 1423.1 This update compares the degree of urbanization rasters (SMOD) of Global Human Settlement Layer between two dataset releases - R2022A (released in 2022) and R2023A (released on 8 May 2023 - https://ghsl.jrc.ec.europa.eu/ghs_smod2023.php) for percentage change in multiple pixel classes and Gini coefficient of Chennai Metropolitan Area urban sprawl bounding box http://bboxfinder.com/#12.439259,79.271851,13.568572,80.351257: #Chennai Metropolitan Area Sprawl Bounding Box 3 - http://bboxfinder.com/#12.439259,79.271851,13.568572,80.351257 - R2022A and R2023A comparison r1data=urban_sprawl_from_raster(79.271851,12.439259,80.351257,13.568572,"testlogs/RemoteSensingGIS/GHS_SMOD_P2030_GLOBE_R2022A_54009_1000_V1_0.tif",dt="Degree of Urbanization R2022A") r2data=urban_sprawl_from_raster(79.271851,12.439259,80.351257,13.568572,"testlogs/RemoteSensingGIS/GHS_SMOD_E2030_GLOBE_R2023A_54009_1000_V1_0.tif",dt="Degree of Urbanization R2023A") compare_raster_data(r1data[0],r2data[0]) 1423.2 Gini index of the bounding has decreased marginally by 0.006% in 1 year between 2022 and 2023 implying less inequality: urbansprawl_gini_coefficient(): Gini Index of the dataset = 0.20324072304374857 (R2022A) urbansprawl_gini_coefficient(): Gini Index of the dataset = 0.20260858566471088 (R2023A) 1423.3 Following are the percentage changes in multiple pixel classes between R2022A and R2023A which according to 4-coloring scheme belong to residential,commercial,industrial and waterbody zones: 1423.3.1 Suburban or Periurban grid cell - Percentage change in 21 : 25.377574370709382 1423.3.2 Low Density Rural grid cell - Percentage change in 12 : -23.575883575883577 1423.3.3 Very Low Density Rural grid cell - Percentage change in 11 : -28.631402183039462 1423.3.4 Urban Center Grid cell - Percentage change in 30 : 26.635514018691588 1423.3.5 Semi Dense Urban cluster grid cell - Percentage change in 22 : -247.22222222222223 1423.3.6 Rural cluster grid cell - Percentage change in 13 : -178.48101265822785 1423.3.7 Dense Urban cluster grid cell - Percentage change in 23 : 20.118343195266274 1423.3.8 Water Grid cell - Percentage change in 10 : -0.7547169811320755 1423.4 Above 1 year trend shows an approximate 20-26% growth in Urban and Periurban areas and corresponding 23-28% decline in rural grid cells implying ~25% urbanization of rural grid cells within the bounding box in 1 year. There is a sharp decline of 178-247% in rural cluster and semi dense urban cluster grid cells of the bounding box within 1 year. It has to be noted Class 21 periurban pixel class expanded by almost 2500% in 3 years between R2019A and R2022A which was analyzed in Section 1313 and 1320 and 1 year growth of Class 21 pixels is 25.37% between 2022 and 2023. 1423.5 GHSL R2023A SMOD dataset and logs have been compressed and committed to repositories at: testlogs/GISUrbanSprawlAnalytics.log.9May2023.xz testlogs/RemoteSensingGIS/GHS_SMOD_E2030_GLOBE_R2023A_54009_1000_V1_0.tif.xz
Srinivasan Kannan
More... over 3 years ago
Section 1422 updated
root
More... over 3 years ago
Updated Section 1422
root
More... over 3 years ago
Section 1422 updated
root
More... over 3 years ago
Updated Requirements.txt
Srinivasan Kannan
More... over 3 years ago
------------------------------------------------------------------------------------------------------------------------------ 1422. (THEORY) Merit of Large Scale Visuals - Homotopy Continuation for Continuous Deformations - integration of PHCPy implementation - 3 May 2023 - related to 2,345,714,715,1203,1232,1233,1236,1237,1240,1287,1288,1417,1418,1421 and all sections on Merit of Large Scale Visuals, Physique Recognition From Quadrilateral mesh isomorphism,Topological Face and Handwriting recognition, Archaeology-Topological Script Recognition, Product Homotopy, Homeomorphism and Topological Deformations, People Analytics, Computational Geometry, Graph Edit Distance and Earth mover distance, Graph Isomorphism (Exact, Approximate and Subgraph),Graph matching,Graph similarity,Image Search,Quadrilateral Mesh Deformation and Isomorphism,Maximum Common Subgraph Problem and Approximate Topological Matching,Text restoration,Polynomial Reconstruction Problem (PRP), Alt's Problem-Four Bar Linkage-Nine Point Synthesis,Unique Identification by Face Recognition-Handwriting-Fingerprint ------------------------------------------------------------------------------------------------------------------------------ 1422.1 PHCPy python interface of PHC has been integrated into HomotopyContinuation.py as separate function def phc_homotopy(start_system) which has been derived and modified from documentation example at http://homepages.math.uic.edu/~jan/phcpy_doc_html/pathtrack.html 1422.2 Function phc_homotopy() accepts a start system of polynomials (p), whose solutions are connected to a target solutions system of polynomials (q) by a homotopy. Homotopy continuation or deformation is by tracking function initialize_standard_tracker() which connects p and q. Functions intialize_standard_solution() and next_standard_solution() keep track of each polynomial defined by increment parameter t. Source and target polynomial systems could be contour polynomials of large scale visual segmentation from 2 images to be matched. 1422.3 Homotopy continuation between p and q are drawn by matplotlib which simulates contour polynomial homotopy in : (1) the human handwriting deformations - same letter written by two individuals are connected by a homotopy continuation (as endpoints are always constant in letter polynomials and only the curves connecting the endpoints are deformed - defined by Product Homotopy shape grammar of 714 and 715) (2) facial deformations in face recognition (3) fingerprint recognition 1422.4 PHCpy has been compiled by target "make phcpy2c3.so" on PHCPack source tree which has been added to repository en masse due to the difficulty in getting a correct PHCPy build. Dependencies are enormous: 1422.4.1 GCC version 7 and GNAT ADA GPL version 7 installed from respective sources - https://packages.debian.org/source/buster/gcc-7 and https://packages.debian.org/buster/gnat-7 1422.4.2 QD cloned and built with -fPIC flag for Quad double datatype - https://github.com/scibuilder/QD 1422.4.3 PHCPy target phcpy2c3.so built with -fPIC flag 1422.5 Homotopy continuation for an example start system are captured in (*) testlogs/HomotopyContinuation.3May2023.jpeg and (*) testlogs/HomotopyContinuation.log.3May2023
Srinivasan Kannan
More... over 3 years ago
------------------------------------------------------------------------------------------------------------------------------ 1422. (THEORY) Merit of Large Scale Visuals - Homotopy Continuation for Continuous Deformations - integration of PHCPy implementation - 3 May 2023 - related to 2,345,714,715,1203,1232,1233,1236,1237,1240,1287,1288,1417,1418,1421 and all sections on Merit of Large Scale Visuals, Physique Recognition From Quadrilateral mesh isomorphism,Topological Face and Handwriting recognition, Archaeology-Topological Script Recognition, Product Homotopy, Homeomorphism and Topological Deformations, People Analytics, Computational Geometry, Graph Edit Distance and Earth mover distance, Graph Isomorphism (Exact, Approximate and Subgraph),Graph matching,Graph similarity,Image Search,Quadrilateral Mesh Deformation and Isomorphism,Maximum Common Subgraph Problem and Approximate Topological Matching,Text restoration,Polynomial Reconstruction Problem (PRP), Alt's Problem-Four Bar Linkage-Nine Point Synthesis,Unique Identification by Face Recognition-Handwriting-Fingerprint ------------------------------------------------------------------------------------------------------------------------------ 1422.1 PHCPy python interface of PHC has been integrated into HomotopyContinuation.py as separate function def phc_homotopy(start_system) which has been derived and modified from documentation example at http://homepages.math.uic.edu/~jan/phcpy_doc_html/pathtrack.html 1422.2 Function phc_homotopy() accepts a start system of polynomials (p), whose solutions are connected to a target solutions system of polynomials (q) by a homotopy. Homotopy continuation or deformation is by tracking function initialize_standard_tracker() which connects p and q. Functions intialize_standard_solution() and next_standard_solution() keep track of each polynomial defined by increment parameter t. Source and target polynomial systems could be contour polynomials of large scale visual segmentation from 2 images to be matched. 1422.3 Homotopy continuation between p and q are drawn by matplotlib which simulates contour polynomial homotopy in : (1) the human handwriting deformations - same letter written by two individuals are connected by a homotopy continuation (as endpoints are always constant in letter polynomials and only the curves connecting the endpoints are deformed - defined by Product Homotopy shape grammar of 714 and 715) (2) facial deformations in face recognition (3) fingerprint recognition 1422.4 PHCpy has been compiled by target "make phcpy2c3.so" on PHCPack source tree which has been added to repository en masse due to the difficulty in getting a correct PHCPy build. Dependencies are enormous: 1422.4.1 GCC version 7 and GNAT ADA GPL version 7 installed from respective sources - https://packages.debian.org/source/buster/gcc-7 and https://packages.debian.org/buster/gnat-7 1422.4.2 QD cloned and built with -fPIC flag for Quad double datatype - https://github.com/scibuilder/QD 1422.4.3 PHCPy target phcpy2c3.so built with -fPIC flag 1422.5 Homotopy continuation for an example start system are captured in (*) testlogs/HomotopyContinuation.3May2023.jpeg and (*) testlogs/HomotopyContinuation.log.3May2023
Srinivasan Kannan
More... over 3 years ago
Updated Requirements.txt
root
More... over 3 years ago
+1421. (THEORY) Merit of Large Scale Visuals - Homotopy Continuation for Continuous Deformations - integration of PHCPack implementation - 1 May 2023 - related to 2,345,714,715,1203,1232,1233,1236,1237,1240,1287,1288,1417,1418 and all sections on Merit of Large Scale Visuals, Physique Recognition From Quadrilateral mesh isomorphism,Topological Face and Handwriting recognition, Archaeology-Topological Script Recognition, Product Homotopy, Homeomorphism and Topological Deformations, People Analytics, Computational Geometry, Graph Edit Distance and Earth mover distance, Graph Isomorphism (Exact, Approximate and Subgraph),Graph matching,Graph similarity,Image Search,Quadrilateral Mesh Deformation and Isomorphism,Maximum Common Subgraph Problem and Approximate Topological Matching,Text restoration,Polynomial Reconstruction Problem (PRP), Alt's Problem-Four Bar Linkage-Nine Point Synthesis,Unique Identification by Face Recognition-Handwriting-Fingerprint^M
root
More... over 3 years ago
------------------------------------------------------------------------------------------------------------------------------ 1421. (THEORY) Merit of Large Scale Visuals - Homotopy Continuation for Continuous Deformations - integration of PHCPack implementation - 1 May 2023 - related to 2,345,714,715,1203,1232,1233,1236,1237,1240,1287,1288,1417,1418 and all sections on Merit of Large Scale Visuals, Physique Recognition From Quadrilateral mesh isomorphism,Topological Face and Handwriting recognition, Archaeology-Topological Script Recognition, Product Homotopy, Homeomorphism and Topological Deformations, People Analytics, Computational Geometry, Graph Edit Distance and Earth mover distance, Graph Isomorphism (Exact, Approximate and Subgraph),Graph matching,Graph similarity,Image Search,Quadrilateral Mesh Deformation and Isomorphism,Maximum Common Subgraph Problem and Approximate Topological Matching,Text restoration,Polynomial Reconstruction Problem (PRP), Alt's Problem-Four Bar Linkage-Nine Point Synthesis,Unique Identification by Face Recognition-Handwriting-Fingerprint ------------------------------------------------------------------------------------------------------------------------------ 1421.1 This update defines a new python source file HomotopyContinuation.py within python-src/HomotopyContinuation that implements a small python function to invoke PHC (Polynomial Homotopy Continuation) commandline by taking a source polynomial systems file as input and writing the continuous deformation solutions to output file. 1421.2 PHC executable can be installed from source in http://homepages.math.uic.edu/~jan/phcpy_doc_html/index.html or https://launchpad.net/ubuntu/jammy/+package/phcpack (Ubuntu apt binaries) 1421.3 Example commandline "phc -p input output" computes homotopy continuation for set of 2 polynomials in file HomotopyContinuation/input: f(x,y)=x*y + 2*y^3 - 3.14E-01; g(x,y)=3*(x + y)*x - 2/3; and writes the solutions and output to HomotopyContinuation/solution and HomotopyContinuation/output by getting user choices through CLI questions. PHC logs are written to python-src/HomotopyContinuation/testlogs/HomotopyContinuation.log.1May2023 by choosing CLI option "2 : intermediate solutions at each step along the paths" 1421.4 Of all Homotopy continuation solutions available, PHC commandline is the simplest and could be useful to simulate homotopy (continuous deformation) of contour polynomials often obtained from images of human handwriting,human facial deformations,fingerprints for unique identification - every feature of human face,handwriting and fingerprint is defineable by a pattern grammar postulated in Sections 714 and 715 and every symbol of the pattern grammar is a contour homotopy. 1421.5 Set of all continuous deformations or homotopy between f(x) and g(x) is parameterized by h(x, t) = (1 − t)f(x) + tg(x) for t in [0,1]. Functions f(x) and g(x) could be two contour polynomials from human face-handwriting-fingerprint or an archaeological inscription. 1421.6 At present PHCPy has not been integrated as phc commandline utility is comprehensive and fast. Unique identification by face-fingerprint-handwriting could be implemented by (#) populating the 2 contour polynomials (from segmentation of two visuals to be compared) within HomotopyContinuation/input (one polynomial from each image) (#) finding the solution polynomials (#) rank the solutions by distance to get best match (lowest Directed Hausdorff distance)
root
More... over 3 years ago
Updated Doxygen Doxyfile config
root
More... over 3 years ago
Updated NeuronRain_Research_Celestial_Pattern_Mining.txt
root
More... over 3 years ago
Section 1418 updated
root
More... over 3 years ago
+1419. (JIRA,THEORY and FEATURE) Computational Geometric Integer Factorization - DMRC PySpark Cloud and Rust High Performance Computing NC versions - Iterative Binary Search implementation - Some bugs fixed and multiple integer benchmarks in Spark and Rust - related to 498,601,1226,1259,1303,1319,1321,1323,1324,1325,1355,1365,1366,1369,1379,1381,1384,1385,1401 and all sections on Integer Factorization (DMRC MapReduce PySpark and NC HPC Rust implementations),Computational Geometry,Set Partitions-to-Tile Cover reduction by Lagranges Four Square Theorem and ILP,Primality Testing,RSA Semiprimes,Almost-Square Bounding Boxes in Large Scale Visual Recognition,Pell's Equation,Brahmagupta's Chakravala,Approximate Factoring by Newton-Raphson Rayshooting queries,Algebraic Geometry of Computational Geometric Factorization,Quantum computation and Derandomization of BQP Factorization by Condorcet Jury Theorem Majority-BQPVoterSAT boolean composition - 25 April 2023
root
More... over 3 years ago
------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- 1419. (JIRA,THEORY and FEATURE) Computational Geometric Integer Factorization - DMRC PySpark Cloud and Rust High Performance Computing NC versions - Iterative Binary Search implementation - Some bugs fixed and multiple integer benchmarks in Spark and Rust - related to 498,601,1226,1259,1303,1319,1321,1323,1324,1325,1355,1365,1366,1369,1379,1381,1384,1385,1401 and all sections on Integer Factorization (DMRC MapReduce PySpark and NC HPC Rust implementations),Computational Geometry,Set Partitions-to-Tile Cover reduction by Lagranges Four Square Theorem and ILP,Primality Testing,RSA Semiprimes,Almost-Square Bounding Boxes in Large Scale Visual Recognition,Pell's Equation,Brahmagupta's Chakravala,Approximate Factoring by Newton-Raphson Rayshooting queries,Algebraic Geometry of Computational Geometric Factorization,Quantum computation and Derandomization of BQP Factorization by Condorcet Jury Theorem Majority-BQPVoterSAT boolean composition - 25 April 2023 ------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- 1419.1 python-src/DiscreteHyperbolicFactorizationUpperbound_TileSearch_Optimized.py has been updated to comment numba JIT imports and decorators as Numba is incompatible with python 3.11. Function def iterative_binary_search_interval_nonpersistent(xl,yl,xr,yr,numfactors) has been updated to cast midpoint computation to int (which was causing floating point values) similar to recursive binary search point location. 1419.2 Factors are added to factors accumulator of Spark after getting string representation of the factor from Decimal by to_eng_string(). 1419.3 python-src/DiscreteHyperbolicFactorizationUpperbound_TileSearch_Optimized_MultipleIntegers.py has been updated to remove str() cast for number_of_factors as "all" factors are computed by default and "all" is a string. 1419.4 100 Consecutive integers from 100000 are factorized by Spark 3.3.0 + Python 3.7.5 and Rust Rayon Parallel Iterator by following commandlines in NeuronRain AstroInfer and NeuronRain Grafit: 1419.4.1 NeuronRain AstroInfer - python3.7 DiscreteHyperbolicFactorizationUpperbound_TileSearch_Optimized_MultipleIntegers.py 100000 100 1 1 2 /home/ksrinivasan/spark-3.3.0-bin-hadoop3/ all 2>&1 > testlogs/DiscreteHyperbolicFactorizationUpperbound_TileSearch_Optimized_MultipleIntegers.log.25April2023 1419.4.2 NeuronRain AstroInfer - python3.7 DiscreteHyperbolicFactorizationUpperbound_TileSearch_Optimized_MultipleIntegers.py 100000 100 1 1 2 /home/ksrinivasan/spark-3.3.0-bin-hadoop3/ all 2>&1 > testlogs/DiscreteHyperbolicFactorizationUpperbound_TileSearch_Optimized_MultipleIntegers.log2.25April2023 1419.4.3 NeuronRain Grafit - cargo run 100000 100 parallel - Programming/Rust/rasterization/src/logs/rasterization.parallel.multipleintegers.log.25April2023 1419.5 PySpark DMRC implementation factorized 100 consecutive integers from 100000 in an average time of 28.768128364 seconds per integer while Rust Rayon Parallel Iterator HPC implementation factorized 100 consecutive integers from 100000 in an average time of ~5.5 milliseconds per integer (or Rust was approximately 5600 times faster per integer on the average, than Spark Python) - Logs-Plots for the benchmarks are available within NeuronRain AsFer repositories: testlogs/DiscreteHyperbolicFactorizationUpperbound_TileSearch_Optimized_MultipleIntegers.jpg testlogs/DiscreteHyperbolicFactorizationUpperbound_TileSearch_Optimized_MultipleIntegers.log.25April2023 (57 integers) testlogs/DiscreteHyperbolicFactorizationUpperbound_TileSearch_Optimized_MultipleIntegers.log2.25April2023 (100 integers) and NeuronRain Grafit repositories: rasterization/src/logs/rasterization.parallel.multipleintegers.log.25April2023 1419.6 This commit resolves JIRA issue https://krishnairesearch.atlassian.net/browse/NEUR-28
root
More... over 3 years ago
Section 1418 updated (2/2)
root
More... over 3 years ago
Section 1418 updated
root
More... over 3 years ago
1418. (THEORY) Merit of Large Scale Visuals - Archaeology - Text Restoration by Product Homotopies of Script Contour Polynomials - 20 April 2023 - related to 2,345,714,1203,1232,1233,1236,1237,1240,1287,1288,1417 and all sections on Merit of Large Scale Visuals, Physique Recognition From Quadrilateral mesh isomorphism,Topological Face and Handwriting recognition, Archaeology-Topological Script Recognition, Product Homotopy, Homeomorphism and Topological Deformations, People Analytics, Computational Geometry, Graph Edit Distance and Earth mover distance, Graph Isomorphism (Exact, Approximate and Subgraph),Graph matching,Graph similarity,Image Search,Quadrilateral Mesh Deformation and Isomorphism,Maximum Common Subgraph Problem and Approximate Topological Matching,Text restoration,Polynomial Reconstruction Problem (PRP)
root
More... over 3 years ago
Section 1236 updated
root
More... over 3 years ago
1417. (JIRA,THEORY and FEATURE) Merit of Large Scale Visuals - Archaeology - Few more updates for drawContours() versus Pyplot and Edges/Contour polynomials of scripts - Tamil and Ashoka Brahmi script example - 17 April 2023 - related to 2,1203,1232,1233,1236,1237,1240,1287,1288 and all sections on Merit of Large Scale Visuals, Physique Recognition From Quadrilateral mesh isomorphism,Topological Face and Handwriting recognition, Archaeology-Topological Script Recognition, Product Homotopy, Homeomorphism and Topological Deformations, People Analytics, Computational Geometry, Graph Edit Distance and Earth mover distance, Graph Isomorphism (Exact, Approximate and Subgraph),Graph matching,Graph similarity,Image Search,Quadrilateral Mesh Deformation and Isomorphism,Maximum Common Subgraph Problem and Approximate Topological Matching,Text restoration
root
More... over 3 years ago
------------------------------------------------------------------------------------------------------------------------------- 1417. (JIRA,THEORY and FEATURE) Merit of Large Scale Visuals - Archaeology - Few more updates for drawContours() versus Pyplot and Edges/Contour polynomials of scripts - Tamil and Ashoka Brahmi script example - 17 April 2023 - related to 2,1203,1232,1233,1236,1237,1240,1287,1288 and all sections on Merit of Large Scale Visuals, Physique Recognition From Quadrilateral mesh isomorphism,Topological Face and Handwriting recognition, Archaeology-Topological Script Recognition, Product Homotopy, Homeomorphism and Topological Deformations, People Analytics, Computational Geometry, Graph Edit Distance and Earth mover distance, Graph Isomorphism (Exact, Approximate and Subgraph),Graph matching,Graph similarity,Image Search,Quadrilateral Mesh Deformation and Isomorphism,Maximum Common Subgraph Problem and Approximate Topological Matching,Text restoration -------------------------------------------------------------------------------------------------------------------------------- 1417.1 Updated python-src/image_pattern_mining/ImageNet/ImageGraph_Keras_Theano.py to introduce another boolean flag usepyplot to choose between pyplot and drawContours() in plotting the contour segmented imagery - There seems to be a problem recently in pyplot and contours are now plotted by drawContours() within image_segmentation_contours(imagefile1,plot="savefig") by default. 1417.2 Either pyplot savefig/show or cv2.imwrite() are invoked based on usepyplot flag to write xxxxx_contoursegmented.jpg (all contour polynomials are drawn). 1417.3 Return tuple of image_segmentation() has been expanded to include cannyedge detected edges. return (ret, img, markers, labels, stats, contourcentroids, facets, triangles, contours, facegraph, edges) 1417.4 An image containing Asokan Brahmi-Tamil Brahmi(Early and Later) inscription and Modern Tamil from https://en.wikipedia.org/wiki/Tamil-Brahmi has been segmented to compute contour polynomials and canny edges which is an algebraic alternative to ingest undeciphered and ancient scripts vis-a-vis contemporary representation of natural language by ASCII-Unicode. Every letter in an undeciphered inscription (e.g Indus-Sarasvati script) is a polynomial curve and storing undeciphered text as huge set of polynomials (equal to number of symbols in the inscription text) opens up possibilities of algebraic and topological analysis (including Voronoi-Delaunay diagrams) of writing systems (For example same symbol in two inscriptions are homeomorphic deformations): seg5=image_segmentation("testlogs/Tamil_brahmi.jpg",voronoi_delaunay=True) 1417.5 Tamil Brahmi and Ashoka Brahmi are derivatives of Brahmi (Tamil Brahmi predates Ashoka Brahmi) which in turn is suspected to be a descendant of Aramaic-Phoenician scripts. Image earlier from https://en.wikipedia.org/wiki/Tamil-Brahmi shows a topological similarity between Tamil Brahmi and Ashoka Brahmi scripts (Letters m and ma in Asokan and Tamil Brahmi having close resemblance are homeomorphic deformations). Polynomial representation of scripts is useful in text restoration - if an ancient text in damaged manuscript has only few fragments of a symbol,missing pieces can be interpolated by contour polynomials learnt from intact symbols in manuscript and accuracy depends on frequency of intact symbols (Section 2.16 - DeepMind Ithaca is 71% accurate in restoring 485 BC Greek text). But the limitation is the storage requirement for polynomials (could be millions, equal to number of symbols). 1417.6 Following are the logs,contour,segment,cannyedge,Voronoi-Delunay images for this update: testlogs/ImageGraph_Keras_Theano.log.17April2023 testlogs/RemoteSensingGIS/Tamil_brahmi_ImageNet_Keras_Theano_Segmentation_FaceGraph.dot testlogs/RemoteSensingGIS/Tamil_brahmi_segmented.jpg testlogs/Tamil_brahmi-cannyedgedetection.jpg testlogs/Tamil_brahmi-contourlabelled.jpg testlogs/Tamil_brahmi-contoursegmented.jpg testlogs/Tamil_brahmi.jpg 1417.7 This commit resolves JIRA issue https://krishnairesearch.atlassian.net/browse/NEUR-27
root
More... over 3 years ago