{ "cells": [ { "cell_type": "markdown", "metadata": {}, "source": [ "# Getting Setup\n", "This section goes through various steps to get setup for outlier detection. For more details see the ingesting ISIS control networks tutorial." ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "## Imports" ] }, { "cell_type": "code", "execution_count": 1, "metadata": { "collapsed": true }, "outputs": [], "source": [ "# Required for autocnet imports\n", "import os\n", "os.environ['ISISROOT'] = '/usgs/cpkgs/anaconda3_linux/envs/isis4.2.0'\n", "\n", "# Autocnet Imports\n", "from autocnet.graph.network import NetworkCandidateGraph\n", "from autocnet.graph.edge import NetworkEdge\n", "from autocnet.io.db.model import Matches, Points\n", "from autocnet.transformation.roi import Roi\n", "\n", "# Helpful Python Modules\n", "import matplotlib.pyplot as plt # plotting package\n", "import numpy as np # numerical computing package" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "## Setting up the NetworkCandidateGraph" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "### Config\n", "\n", "The config various settings that autocnet will use when connecting to other services. Primarily, the config is used to define:\n", "\n", "- The database your NetworkCandidateGraph will use\n", "- The redis queue and slurm settings for cluster based processing\n", "- The spatial reference system for geometries such as image footprints" ] }, { "cell_type": "code", "execution_count": 2, "metadata": { "collapsed": true }, "outputs": [], "source": [ "config_path = 'config.yml'" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "### Loading the control network\n", "This cell will check if the database your config file points to already has a control network ingested in it. If it doesn't then it goes through the steps from the ingesting ISIS control networks tutorial." ] }, { "cell_type": "code", "execution_count": 3, "metadata": { "scrolled": true }, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "Network already in database\n" ] }, { "name": "stderr", "output_type": "stream", "text": [ "/work/users/jmapel/anaconda_local/envs/autocnet_local/lib/python3.7/site-packages/sqlalchemy/orm/relationships.py:1997: SAWarning: Setting backref / back_populates on relationship Overlay.points to refer to viewonly relationship Points.overlay should include sync_backref=False set on the Overlay.points relationship. (this warning may be suppressed after 10 occurrences)\n", " (rel_b, rel_a, rel_b),\n", "/work/users/jmapel/anaconda_local/envs/autocnet_local/lib/python3.7/site-packages/sqlalchemy/orm/relationships.py:1997: SAWarning: Setting backref / back_populates on relationship Points.overlay to refer to viewonly relationship Overlay.points should include sync_backref=False set on the Points.overlay relationship. (this warning may be suppressed after 10 occurrences)\n", " (rel_b, rel_a, rel_b),\n" ] } ], "source": [ "data_directory = \"/work/projects/control_network_metrics/tutorials/isis_ingestion\"\n", "lis_path = os.path.join(data_directory, \"apollo_lronac_cubes.lis\")\n", "cnet_path = os.path.join(data_directory, \"AS15_landingsite_apollolro_jig1.net\")\n", "\n", "ncg = NetworkCandidateGraph()\n", "ncg.config_from_file(config_path)\n", "ncg.from_database()\n", "if len(ncg) == 0:\n", " print(f'Ingesting control network {cnet_path}.')\n", " ncg = NetworkCandidateGraph.from_cnet(cnet_path, lis_path, config_path)\n", "else:\n", " print('Network already in database')" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "### Check network\n", "Look at the graph of the network to ensure it isn't malformed. Each image in the network is represented by a node and overlapping images have an edge between their nodes. We will be doing pair-wise outlier detection, so we will check each edge in the network graph" ] }, { "cell_type": "code", "execution_count": 4, "metadata": { "scrolled": true }, "outputs": [ { "data": { "text/plain": [ "" ] }, "execution_count": 4, "metadata": {}, "output_type": "execute_result" }, { "data": { "image/png": "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\n", "text/plain": [ "
" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "ncg.plot()" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "### Getting the pairwise image matches\n", "The first process we need to do is collect all of the pairwise matches between images. Our control network currently contains control points and all of the measurements of them. We need to convert these multi-image relationships into all of the common points between each pair of images.\n", "\n", "To do this, we're going to use Autocnet's apply function to run parallel processing on each edge of the NetworkCandidateGraph" ] }, { "cell_type": "code", "execution_count": 5, "metadata": { "collapsed": true }, "outputs": [], "source": [ "?NetworkCandidateGraph.apply" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "#### SLURM parameters\n", "These parameters will be used when creating SLURM jobs for cluster processing via the apply function. Depending on the complexity of the jobs you are running, you may want to change the walltime and arraychunk parameters." ] }, { "cell_type": "code", "execution_count": 6, "metadata": { "collapsed": true }, "outputs": [], "source": [ "walltime=\"00:30:00\"\n", "log_dir = '/scratch/jmapel/autocnet_tut/logs'\n", "arraychunk=75\n", "chunksize=16723" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "#### Convert control measures and points to image matches\n", "This cell uses apply to run the network_to_matches function on each edge. It is very important that this function only get run once per edge or it will add duplicate matches. So, this cell also contains a check that skips the function if the database already has matches in it." ] }, { "cell_type": "code", "execution_count": 7, "metadata": { "collapsed": true }, "outputs": [], "source": [ "?NetworkEdge.network_to_matches" ] }, { "cell_type": "code", "execution_count": 8, "metadata": { "scrolled": true }, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "Loading matches table\n" ] } ], "source": [ "with ncg.session_scope() as session:\n", " num_matches = session.query(Matches).count()\n", "if num_matches == 0:\n", " print(\"Loading matches table\")\n", " njobs = ncg.apply('network_to_matches', \n", " on='edges',\n", " # SLURM kwargs\n", " walltime=walltime,\n", " log_dir=os.path.join(log_dir, 'matches'),\n", " arraychunk=arraychunk,\n", " chunksize=chunksize)\n", "else:\n", " print(\"Matches table already populated\")" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "### Looking at the matches\n", "The pairwise image matches are stored on the edges of the graph and can be accessed via the networkX graph or the matches table in the database" ] }, { "cell_type": "code", "execution_count": 9, "metadata": { "scrolled": true }, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "Edge (10, 6) has 440 matches\n", "Edge (10, 3) has 4 matches\n", "Edge (10, 4) has 4 matches\n", "Edge (10, 2) has 4 matches\n", "Edge (10, 11) has 39 matches\n", "Edge (10, 8) has 477 matches\n", "Edge (10, 5) has 4 matches\n", "Edge (10, 7) has 54 matches\n", "Edge (1, 3) has 178 matches\n", "Edge (1, 4) has 82 matches\n", "Edge (1, 2) has 197 matches\n", "Edge (1, 5) has 19 matches\n", "Edge (6, 3) has 4 matches\n", "Edge (6, 4) has 4 matches\n", "Edge (6, 2) has 4 matches\n", "Edge (6, 11) has 38 matches\n", "Edge (6, 8) has 419 matches\n", "Edge (6, 5) has 4 matches\n", "Edge (6, 7) has 46 matches\n", "Edge (3, 9) has 3 matches\n", "Edge (3, 4) has 271 matches\n", "Edge (3, 2) has 274 matches\n", "Edge (3, 11) has 5 matches\n", "Edge (3, 8) has 4 matches\n", "Edge (3, 5) has 181 matches\n", "Edge (3, 7) has 2 matches\n", "Edge (9, 4) has 3 matches\n", "Edge (9, 2) has 3 matches\n", "Edge (9, 11) has 239 matches\n", "Edge (9, 8) has 42 matches\n", "Edge (9, 5) has 3 matches\n", "Edge (9, 7) has 213 matches\n", "Edge (4, 2) has 166 matches\n", "Edge (4, 11) has 5 matches\n", "Edge (4, 8) has 4 matches\n", "Edge (4, 5) has 199 matches\n", "Edge (4, 7) has 2 matches\n", "Edge (2, 11) has 5 matches\n", "Edge (2, 8) has 4 matches\n", "Edge (2, 5) has 89 matches\n", "Edge (2, 7) has 2 matches\n", "Edge (11, 8) has 171 matches\n", "Edge (11, 5) has 5 matches\n", "Edge (11, 7) has 275 matches\n", "Edge (8, 5) has 4 matches\n", "Edge (8, 7) has 184 matches\n", "Edge (5, 7) has 2 matches\n" ] } ], "source": [ "for source, dest, edge in ncg.edges(data='data'):\n", " print(f'Edge ({source}, {dest}) has {len(edge.matches)} matches')" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "## Outlier Detection" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "### Reprojective Error\n", "These two functions check for outliers by attempting to reproject measures between images. Any pair of measures that do not repoject to each other wth in a given tolerance are flagged." ] }, { "cell_type": "code", "execution_count": null, "metadata": { "collapsed": true }, "outputs": [], "source": [ "?NetworkEdge.compute_fundamental_matrix" ] }, { "cell_type": "code", "execution_count": null, "metadata": { "collapsed": true }, "outputs": [], "source": [ "?NetworkEdge.compute_homography" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "These cells also demonstrate how to pass kwargs through apply. Any kwargs that are not specific to the apply function are passed on to the function being applied." ] }, { "cell_type": "code", "execution_count": 10, "metadata": { "collapsed": true, "scrolled": true }, "outputs": [], "source": [ "njobs = ncg.apply('compute_fundamental_matrix', \n", " on='edges',\n", " # homography kwargs\n", " method='mle',\n", " reproj_threshold=5,\n", " # SLURM kwargs\n", " walltime=walltime,\n", " log_dir=os.path.join(log_dir, 'fundamental'),\n", " arraychunk=arraychunk,\n", " chunksize=chunksize)" ] }, { "cell_type": "code", "execution_count": 11, "metadata": { "collapsed": true }, "outputs": [], "source": [ "njobs = ncg.apply('compute_homography', \n", " on='edges',\n", " # fundamental matrix kwargs\n", " method='lmeds',\n", " reproj_threshold=5,\n", " # SLURM kwargs\n", " walltime=walltime,\n", " log_dir=os.path.join(log_dir, 'homography'),\n", " arraychunk=arraychunk,\n", " chunksize=chunksize)" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "You can check the queue_length property on the NetworkCandidateGraph object to see how many jobs are either waiting to be processed or in process. You can also use the squeue command on a command line or in your notebook to check what jobs slurm has." ] }, { "cell_type": "code", "execution_count": 14, "metadata": { "scrolled": false }, "outputs": [ { "data": { "text/plain": [ "0" ] }, "execution_count": 14, "metadata": {}, "output_type": "execute_result" } ], "source": [ "ncg.queue_length" ] }, { "cell_type": "code", "execution_count": 13, "metadata": { "scrolled": false }, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ " JOBID PARTITION NAME USER ST TIME NODES NODELIST(REASON) \r\n", " 17780111 longall jupyter- jmapel R 2:09:58 1 neb16 \r\n", " 17796596_1 shortall AutoCNet jmapel R 0:03 1 neb14 \r\n", " 17796596_2 shortall AutoCNet jmapel R 0:03 1 neb14 \r\n", " 17796596_3 shortall AutoCNet jmapel R 0:03 1 neb14 \r\n", " 17796596_4 shortall AutoCNet jmapel R 0:03 1 neb14 \r\n", " 17796596_5 shortall AutoCNet jmapel R 0:03 1 neb14 \r\n", " 17796596_6 shortall AutoCNet jmapel R 0:03 1 neb14 \r\n", " 17796596_7 shortall AutoCNet jmapel R 0:03 1 neb14 \r\n", " 17796596_8 shortall AutoCNet jmapel R 0:03 1 neb14 \r\n", " 17796596_9 shortall AutoCNet jmapel R 0:03 1 neb14 \r\n", " 17796596_10 shortall AutoCNet jmapel R 0:03 1 neb15 \r\n", " 17796596_11 shortall AutoCNet jmapel R 0:03 1 neb15 \r\n", " 17796596_12 shortall AutoCNet jmapel R 0:03 1 neb15 \r\n", " 17796596_13 shortall AutoCNet jmapel R 0:03 1 neb15 \r\n", " 17796596_14 shortall AutoCNet jmapel R 0:03 1 neb15 \r\n", " 17796596_15 shortall AutoCNet jmapel R 0:03 1 neb15 \r\n", " 17796596_16 shortall AutoCNet jmapel R 0:03 1 neb15 \r\n", " 17796596_17 shortall AutoCNet jmapel R 0:03 1 neb15 \r\n", " 17796596_18 shortall AutoCNet jmapel R 0:03 1 neb15 \r\n", " 17796596_19 shortall AutoCNet jmapel R 0:03 1 neb15 \r\n", " 17796596_20 shortall AutoCNet jmapel R 0:03 1 neb15 \r\n", " 17796596_21 shortall AutoCNet jmapel R 0:03 1 neb15 \r\n", " 17796596_22 shortall AutoCNet jmapel R 0:03 1 neb15 \r\n", " 17796596_23 shortall AutoCNet jmapel R 0:03 1 neb15 \r\n", " 17796596_24 shortall AutoCNet jmapel R 0:03 1 neb15 \r\n", " 17796596_25 shortall AutoCNet jmapel R 0:03 1 neb15 \r\n", " 17796596_26 shortall AutoCNet jmapel R 0:03 1 neb15 \r\n", " 17796596_27 shortall AutoCNet jmapel R 0:03 1 neb15 \r\n", " 17796596_28 shortall AutoCNet jmapel R 0:03 1 neb16 \r\n", " 17796596_29 shortall AutoCNet jmapel R 0:03 1 neb16 \r\n", " 17796596_30 shortall AutoCNet jmapel R 0:03 1 neb16 \r\n", " 17796596_31 shortall AutoCNet jmapel R 0:03 1 neb16 \r\n", " 17796596_32 shortall AutoCNet jmapel R 0:03 1 neb16 \r\n", " 17796596_33 shortall AutoCNet jmapel R 0:03 1 neb16 \r\n", " 17796596_34 shortall AutoCNet jmapel R 0:03 1 neb16 \r\n", " 17796596_35 shortall AutoCNet jmapel R 0:03 1 neb16 \r\n", " 17796596_36 shortall AutoCNet jmapel R 0:03 1 neb16 \r\n", " 17796596_37 shortall AutoCNet jmapel R 0:03 1 neb16 \r\n", " 17796596_38 shortall AutoCNet jmapel R 0:03 1 neb16 \r\n", " 17796596_39 shortall AutoCNet jmapel R 0:03 1 neb16 \r\n", " 17796596_40 shortall AutoCNet jmapel R 0:03 1 neb16 \r\n", " 17796596_41 shortall AutoCNet jmapel R 0:03 1 neb16 \r\n", " 17796596_42 shortall AutoCNet jmapel R 0:03 1 neb16 \r\n", " 17796596_43 shortall AutoCNet jmapel R 0:03 1 neb16 \r\n", " 17796596_44 shortall AutoCNet jmapel R 0:03 1 neb16 \r\n", " 17796596_45 shortall AutoCNet jmapel R 0:03 1 neb16 \r\n", " 17796596_46 shortall AutoCNet jmapel R 0:03 1 neb16 \r\n", " 17796596_47 shortall AutoCNet jmapel R 0:03 1 neb16 \r\n" ] } ], "source": [ "!squeue -u jmapel" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "### Looking at the results\n", "The reprojective error checks add a property to each called masks. This is a Pandas dataframe that contains a column for each check that has been done on the edge. If the row for a match has a true in it, then that match passed the column's check. Conversely, if the row for a match has a false in it, then that match failsed the column's check. We can use some dataframe techniques to look at our results" ] }, { "cell_type": "code", "execution_count": 15, "metadata": { "scrolled": true }, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "edge (10, 6)\n", "num matches: 440\n", "passed homography: 440\n", "passed fundamental: 426\n", "too few matches to compute fundamental matrix\n", "\n", "edge (10, 11)\n", "num matches: 39\n", "passed homography: 30\n", "passed fundamental: 38\n", "too few matches to compute fundamental matrix\n", "\n", "edge (10, 8)\n", "num matches: 477\n", "passed homography: 369\n", "passed fundamental: 447\n", "too few matches to compute fundamental matrix\n", "\n", "edge (10, 7)\n", "num matches: 54\n", "passed homography: 45\n", "passed fundamental: 54\n", "too few matches to compute fundamental matrix\n", "\n", "edge (1, 3)\n", "num matches: 178\n", "passed homography: 138\n", "passed fundamental: 91\n", "too few matches to compute fundamental matrix\n", "\n", "edge (1, 4)\n", "num matches: 82\n", "passed homography: 61\n", "passed fundamental: 52\n", "too few matches to compute fundamental matrix\n", "\n", "edge (1, 2)\n", "num matches: 197\n", "passed homography: 152\n", "passed fundamental: 134\n", "too few matches to compute fundamental matrix\n", "\n", "edge (1, 5)\n", "num matches: 19\n", "passed homography: 17\n", "passed fundamental: 16\n", "too few matches to compute fundamental matrix\n", "\n", "edge (6, 11)\n", "num matches: 38\n", "passed homography: 37\n", "passed fundamental: 38\n", "too few matches to compute fundamental matrix\n", "\n", "edge (6, 8)\n", "num matches: 419\n", "passed homography: 419\n", "passed fundamental: 393\n", "too few matches to compute fundamental matrix\n", "\n", "edge (6, 7)\n", "num matches: 46\n", "passed homography: 39\n", "passed fundamental: 45\n", "too few matches to compute fundamental matrix\n", "\n", "edge (3, 4)\n", "num matches: 271\n", "passed homography: 255\n", "passed fundamental: 177\n", "too few matches to compute fundamental matrix\n", "\n", "edge (3, 2)\n", "num matches: 274\n", "passed homography: 237\n", "passed fundamental: 174\n", "too few matches to compute fundamental matrix\n", "\n", "edge (3, 5)\n", "num matches: 181\n", "passed homography: 162\n", "passed fundamental: 93\n", "too few matches to compute fundamental matrix\n", "\n", "edge (9, 11)\n", "num matches: 239\n", "passed homography: 154\n", "passed fundamental: 193\n", "too few matches to compute fundamental matrix\n", "\n", "edge (9, 8)\n", "num matches: 42\n", "passed homography: 38\n", "passed fundamental: 39\n", "too few matches to compute fundamental matrix\n", "\n", "edge (9, 7)\n", "num matches: 213\n", "passed homography: 138\n", "passed fundamental: 191\n", "too few matches to compute fundamental matrix\n", "\n", "edge (4, 2)\n", "num matches: 166\n", "passed homography: 157\n", "passed fundamental: 98\n", "too few matches to compute fundamental matrix\n", "\n", "edge (4, 5)\n", "num matches: 199\n", "passed homography: 169\n", "passed fundamental: 133\n", "too few matches to compute fundamental matrix\n", "\n", "edge (2, 5)\n", "num matches: 89\n", "passed homography: 81\n", "passed fundamental: 62\n", "too few matches to compute fundamental matrix\n", "\n", "edge (11, 8)\n", "num matches: 171\n", "passed homography: 171\n", "passed fundamental: 154\n", "too few matches to compute fundamental matrix\n", "\n", "edge (11, 7)\n", "num matches: 275\n", "passed homography: 190\n", "passed fundamental: 272\n", "too few matches to compute fundamental matrix\n", "\n", "edge (8, 7)\n", "num matches: 184\n", "passed homography: 117\n", "passed fundamental: 167\n", "too few matches to compute fundamental matrix\n", "\n" ] } ], "source": [ "for source, dest, edge in ncg.edges(data=\"data\"):\n", " num_matches = len(edge.matches)\n", " # MLE requires at least 8 points so skip anything with too few\n", " if num_matches < 8:\n", " continue\n", " print(f'edge ({source}, {dest})')\n", " print('num matches:', num_matches)\n", " print('passed homography:', sum(edge.masks['homography']))\n", " print('passed fundamental:', sum(edge.masks['fundamental']))\n", " print('too few matches to compute fundamental matrix')\n", " print('')" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "## Digging in on edge (9, 7)\n", "Edge (9, 7) has a lot of matches but many failures; let's take a closer look at it." ] }, { "cell_type": "code", "execution_count": 16, "metadata": { "scrolled": false }, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "\n", " Source Image Index: 7\n", " Destination Image Index: 9\n", " Available Masks: fundamental homography\n", "match_id \n", "1171 True False\n", "1172 True False\n", "1173 True False\n", "1174 True False\n", "1175 True False\n", "... ... ...\n", "1455 True True\n", "1457 True True\n", "1458 True True\n", "1460 True True\n", "1462 True False\n", "\n", "[213 rows x 2 columns]\n", " \n", "{'data': \n", " NodeID: 9\n", " Image Name: /work/projects/control_network_metrics/tutorials/isis_ingestion/M102135625RE.lev1_8b.cub\n", " Image PATH: /work/projects/control_network_metrics/tutorials/isis_ingestion/M102135625RE.lev1_8b.cub\n", " Number Keypoints: 0\n", " Available Masks : Empty DataFrame\n", "Columns: []\n", "Index: []\n", " Type: \n", " }\n", "{'data': \n", " NodeID: 7\n", " Image Name: /work/projects/control_network_metrics/tutorials/isis_ingestion/M102128467RE.lev1_8b.cub\n", " Image PATH: /work/projects/control_network_metrics/tutorials/isis_ingestion/M102128467RE.lev1_8b.cub\n", " Number Keypoints: 0\n", " Available Masks : Empty DataFrame\n", "Columns: []\n", "Index: []\n", " Type: \n", " }\n" ] } ], "source": [ "edge_9_7 = ncg.edges[(9, 7)]['data']\n", "image_9 = ncg.nodes[9]\n", "image_7 = ncg.nodes[7]\n", "print(edge_9_7)\n", "print(image_9)\n", "print(image_7)" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "### Looking at the matches that failed outlier detection\n", "We can use the masks dataframe to index the matches dataframe on our edge and see the matches that failed each check" ] }, { "cell_type": "code", "execution_count": 17, "metadata": { "scrolled": true }, "outputs": [ { "data": { "text/html": [ "
\n", "\n", "\n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", "
geompoint_idsource_measure_iddestin_measure_idsourcesource_idxdestinationdestination_idxlatlon...source_apriori_xsource_apriori_ydestination_xdestination_ydestination_apriori_xdestination_apriori_yshift_xshift_yoriginal_destination_xoriginal_destination_y
id
1178None788232423257090NoneNone...2231.1216213.7180.31616595.5204.20516599.5NoneNoneNoneNone
1228None1269367836777090NoneNone...3082.246807.211001.127160.22978.3517163.4NoneNoneNoneNone
1230None1271368436837090NoneNone...3908.226813.231773.387180.671797.617184.37NoneNoneNoneNone
1214None1243361036097090NoneNone...3983.411699.581877.612095.961840.912099.48NoneNoneNoneNone
1222None1261365336527090NoneNone...2674.815831.07493.56213.5-0.5-0.5NoneNoneNoneNone
1223None1262365536547090NoneNone...3078.795826.54893.0476201.25918.5376205.46NoneNoneNoneNone
1238None1283371637157090NoneNone...2557.448894.67398.2339273.58424.9719277.79NoneNoneNoneNone
1246None1296375037497090NoneNone...3361.97109211318.8911293.31356.3311296.8NoneNoneNoneNone
1247None1297375337527090NoneNone...4664.1810913.22828.511267.72791.5111270.5NoneNoneNoneNone
1250None1304376937687090NoneNone...2488.5112971.5411.56513351.7429.92413355.6NoneNoneNoneNone
1251None1305377237717090NoneNone...2955.1612943.51011.0613290.9991.05113294NoneNoneNoneNone
1252None1306377437737090NoneNone...3363.3812945.51387.8913317.51405.8813321.7NoneNoneNoneNone
1256None1318380238017090NoneNone...2520.4315001.8471.54415381495.31515385NoneNoneNoneNone
1264None1328382938287090NoneNone...4237.1716023.82321.6716425.32309.8916428.2NoneNoneNoneNone
1267None1333384138437090NoneNone...3323.35170141364.517390.5-0.5-0.5NoneNoneNoneNone
1269None1335384738497090NoneNone...419517097.32265.517466.5-0.5-0.5NoneNoneNoneNone
1278None1341386538647090NoneNone...3774.6818065.91926.8218441.71900.7918444.7NoneNoneNoneNone
1280None1345387538747090NoneNone...2409.2519093414.64519472.6439.2519476.8NoneNoneNoneNone
1312None1380397739767090NoneNone...4117.9124208.42376.63246042353.0124607.2NoneNoneNoneNone
1324None1392401140107090NoneNone...3231.9826258.61420.526633.5-0.5-0.5NoneNoneNoneNone
1325None1393401440137090NoneNone...3602.8926266.71781.3326634.11807.0326638.5NoneNoneNoneNone
1334None1398402840277090NoneNone...3667.7827259.91867.9427626.61892.4727630.9NoneNoneNoneNone
\n", "

22 rows × 22 columns

\n", "
" ], "text/plain": [ " geom point_id source_measure_id destin_measure_id source source_idx \\\n", "id \n", "1178 None 788 2324 2325 7 0 \n", "1228 None 1269 3678 3677 7 0 \n", "1230 None 1271 3684 3683 7 0 \n", "1214 None 1243 3610 3609 7 0 \n", "1222 None 1261 3653 3652 7 0 \n", "1223 None 1262 3655 3654 7 0 \n", "1238 None 1283 3716 3715 7 0 \n", "1246 None 1296 3750 3749 7 0 \n", "1247 None 1297 3753 3752 7 0 \n", "1250 None 1304 3769 3768 7 0 \n", "1251 None 1305 3772 3771 7 0 \n", "1252 None 1306 3774 3773 7 0 \n", "1256 None 1318 3802 3801 7 0 \n", "1264 None 1328 3829 3828 7 0 \n", "1267 None 1333 3841 3843 7 0 \n", "1269 None 1335 3847 3849 7 0 \n", "1278 None 1341 3865 3864 7 0 \n", "1280 None 1345 3875 3874 7 0 \n", "1312 None 1380 3977 3976 7 0 \n", "1324 None 1392 4011 4010 7 0 \n", "1325 None 1393 4014 4013 7 0 \n", "1334 None 1398 4028 4027 7 0 \n", "\n", " destination destination_idx lat lon ... source_apriori_x \\\n", "id ... \n", "1178 9 0 None None ... 2231.12 \n", "1228 9 0 None None ... 3082.24 \n", "1230 9 0 None None ... 3908.22 \n", "1214 9 0 None None ... 3983.41 \n", "1222 9 0 None None ... 2674.81 \n", "1223 9 0 None None ... 3078.79 \n", "1238 9 0 None None ... 2557.44 \n", "1246 9 0 None None ... 3361.97 \n", "1247 9 0 None None ... 4664.18 \n", "1250 9 0 None None ... 2488.51 \n", "1251 9 0 None None ... 2955.16 \n", "1252 9 0 None None ... 3363.38 \n", "1256 9 0 None None ... 2520.43 \n", "1264 9 0 None None ... 4237.17 \n", "1267 9 0 None None ... 3323.35 \n", "1269 9 0 None None ... 4195 \n", "1278 9 0 None None ... 3774.68 \n", "1280 9 0 None None ... 2409.25 \n", "1312 9 0 None None ... 4117.91 \n", "1324 9 0 None None ... 3231.98 \n", "1325 9 0 None None ... 3602.89 \n", "1334 9 0 None None ... 3667.78 \n", "\n", " source_apriori_y destination_x destination_y destination_apriori_x \\\n", "id \n", "1178 16213.7 180.316 16595.5 204.205 \n", "1228 6807.21 1001.12 7160.22 978.351 \n", "1230 6813.23 1773.38 7180.67 1797.61 \n", "1214 1699.58 1877.61 2095.96 1840.91 \n", "1222 5831.07 493.5 6213.5 -0.5 \n", "1223 5826.54 893.047 6201.25 918.537 \n", "1238 8894.67 398.233 9273.58 424.971 \n", "1246 10921 1318.89 11293.3 1356.33 \n", "1247 10913.2 2828.5 11267.7 2791.51 \n", "1250 12971.5 411.565 13351.7 429.924 \n", "1251 12943.5 1011.06 13290.9 991.051 \n", "1252 12945.5 1387.89 13317.5 1405.88 \n", "1256 15001.8 471.544 15381 495.315 \n", "1264 16023.8 2321.67 16425.3 2309.89 \n", "1267 17014 1364.5 17390.5 -0.5 \n", "1269 17097.3 2265.5 17466.5 -0.5 \n", "1278 18065.9 1926.82 18441.7 1900.79 \n", "1280 19093 414.645 19472.6 439.25 \n", "1312 24208.4 2376.63 24604 2353.01 \n", "1324 26258.6 1420.5 26633.5 -0.5 \n", "1325 26266.7 1781.33 26634.1 1807.03 \n", "1334 27259.9 1867.94 27626.6 1892.47 \n", "\n", " destination_apriori_y shift_x shift_y original_destination_x \\\n", "id \n", "1178 16599.5 None None None \n", "1228 7163.4 None None None \n", "1230 7184.37 None None None \n", "1214 2099.48 None None None \n", "1222 -0.5 None None None \n", "1223 6205.46 None None None \n", "1238 9277.79 None None None \n", "1246 11296.8 None None None \n", "1247 11270.5 None None None \n", "1250 13355.6 None None None \n", "1251 13294 None None None \n", "1252 13321.7 None None None \n", "1256 15385 None None None \n", "1264 16428.2 None None None \n", "1267 -0.5 None None None \n", "1269 -0.5 None None None \n", "1278 18444.7 None None None \n", "1280 19476.8 None None None \n", "1312 24607.2 None None None \n", "1324 -0.5 None None None \n", "1325 26638.5 None None None \n", "1334 27630.9 None None None \n", "\n", " original_destination_y \n", "id \n", "1178 None \n", "1228 None \n", "1230 None \n", "1214 None \n", "1222 None \n", "1223 None \n", "1238 None \n", "1246 None \n", "1247 None \n", "1250 None \n", "1251 None \n", "1252 None \n", "1256 None \n", "1264 None \n", "1267 None \n", "1269 None \n", "1278 None \n", "1280 None \n", "1312 None \n", "1324 None \n", "1325 None \n", "1334 None \n", "\n", "[22 rows x 22 columns]" ] }, "execution_count": 17, "metadata": {}, "output_type": "execute_result" } ], "source": [ "failed_fundamental = edge_9_7.matches.loc[(~edge_9_7.masks[['fundamental']]).all(axis=1)]\n", "failed_fundamental" ] }, { "cell_type": "code", "execution_count": 18, "metadata": { "scrolled": true }, "outputs": [ { "data": { "text/html": [ "
\n", "\n", "\n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", "
geompoint_idsource_measure_iddestin_measure_idsourcesource_idxdestinationdestination_idxlatlon...source_apriori_xsource_apriori_ydestination_xdestination_ydestination_apriori_xdestination_apriori_yshift_xshift_yoriginal_destination_xoriginal_destination_y
id
1177None787232123227090NoneNone...2193.7215099.1124.43115480.9148.70315485.1NoneNoneNoneNone
1180None790233123327090NoneNone...2058.3321232.276.91321614.3101.83121618.5NoneNoneNoneNone
1181None791233523367090NoneNone...2055.9922259.590.960322641.3115.84722645.5NoneNoneNoneNone
1182None792233823397090NoneNone...2077.7123322.8233.41923718.5208.58823721.5NoneNoneNoneNone
1358None1416407340727090NoneNone...4388.9537514.93240.27378423225.5137843.1NoneNoneNoneNone
..................................................................
1288None1354390139007090NoneNone...4633.2620103.52761.520468.5-0.5-0.5NoneNoneNoneNone
1289None1356390639057090NoneNone...2449.7421087.8490.16921466.9514.35421471NoneNoneNoneNone
1290None1357390939087090NoneNone...2893.7921162.4958.41221537.6983.27621541.8NoneNoneNoneNone
1291None1358391239117090NoneNone...3285.8321131.41388.3921530.81364.0221533.8NoneNoneNoneNone
1292None1359391539147090NoneNone...3685.2521185.81808.521558.5-0.5-0.5NoneNoneNoneNone
\n", "

75 rows × 22 columns

\n", "
" ], "text/plain": [ " geom point_id source_measure_id destin_measure_id source source_idx \\\n", "id \n", "1177 None 787 2321 2322 7 0 \n", "1180 None 790 2331 2332 7 0 \n", "1181 None 791 2335 2336 7 0 \n", "1182 None 792 2338 2339 7 0 \n", "1358 None 1416 4073 4072 7 0 \n", "... ... ... ... ... ... ... \n", "1288 None 1354 3901 3900 7 0 \n", "1289 None 1356 3906 3905 7 0 \n", "1290 None 1357 3909 3908 7 0 \n", "1291 None 1358 3912 3911 7 0 \n", "1292 None 1359 3915 3914 7 0 \n", "\n", " destination destination_idx lat lon ... source_apriori_x \\\n", "id ... \n", "1177 9 0 None None ... 2193.72 \n", "1180 9 0 None None ... 2058.33 \n", "1181 9 0 None None ... 2055.99 \n", "1182 9 0 None None ... 2077.71 \n", "1358 9 0 None None ... 4388.95 \n", "... ... ... ... ... ... ... \n", "1288 9 0 None None ... 4633.26 \n", "1289 9 0 None None ... 2449.74 \n", "1290 9 0 None None ... 2893.79 \n", "1291 9 0 None None ... 3285.83 \n", "1292 9 0 None None ... 3685.25 \n", "\n", " source_apriori_y destination_x destination_y destination_apriori_x \\\n", "id \n", "1177 15099.1 124.431 15480.9 148.703 \n", "1180 21232.2 76.913 21614.3 101.831 \n", "1181 22259.5 90.9603 22641.3 115.847 \n", "1182 23322.8 233.419 23718.5 208.588 \n", "1358 37514.9 3240.27 37842 3225.51 \n", "... ... ... ... ... \n", "1288 20103.5 2761.5 20468.5 -0.5 \n", "1289 21087.8 490.169 21466.9 514.354 \n", "1290 21162.4 958.412 21537.6 983.276 \n", "1291 21131.4 1388.39 21530.8 1364.02 \n", "1292 21185.8 1808.5 21558.5 -0.5 \n", "\n", " destination_apriori_y shift_x shift_y original_destination_x \\\n", "id \n", "1177 15485.1 None None None \n", "1180 21618.5 None None None \n", "1181 22645.5 None None None \n", "1182 23721.5 None None None \n", "1358 37843.1 None None None \n", "... ... ... ... ... \n", "1288 -0.5 None None None \n", "1289 21471 None None None \n", "1290 21541.8 None None None \n", "1291 21533.8 None None None \n", "1292 -0.5 None None None \n", "\n", " original_destination_y \n", "id \n", "1177 None \n", "1180 None \n", "1181 None \n", "1182 None \n", "1358 None \n", "... ... \n", "1288 None \n", "1289 None \n", "1290 None \n", "1291 None \n", "1292 None \n", "\n", "[75 rows x 22 columns]" ] }, "execution_count": 18, "metadata": {}, "output_type": "execute_result" } ], "source": [ "failed_homography = edge_9_7.matches.loc[(~edge_9_7.masks[['homography']]).all(axis=1)]\n", "failed_homography" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "We can even use some techniques to see the matches that failed both checks" ] }, { "cell_type": "code", "execution_count": 19, "metadata": {}, "outputs": [ { "data": { "text/html": [ "
\n", "\n", "\n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", "
geompoint_idsource_measure_iddestin_measure_idsourcesource_idxdestinationdestination_idxlatlon...source_apriori_xsource_apriori_ydestination_xdestination_ydestination_apriori_xdestination_apriori_yshift_xshift_yoriginal_destination_xoriginal_destination_y
id
1178None788232423257090NoneNone...2231.1216213.7180.31616595.5204.20516599.5NoneNoneNoneNone
1228None1269367836777090NoneNone...3082.246807.211001.127160.22978.3517163.4NoneNoneNoneNone
1230None1271368436837090NoneNone...3908.226813.231773.387180.671797.617184.37NoneNoneNoneNone
1214None1243361036097090NoneNone...3983.411699.581877.612095.961840.912099.48NoneNoneNoneNone
1222None1261365336527090NoneNone...2674.815831.07493.56213.5-0.5-0.5NoneNoneNoneNone
1223None1262365536547090NoneNone...3078.795826.54893.0476201.25918.5376205.46NoneNoneNoneNone
1264None1328382938287090NoneNone...4237.1716023.82321.6716425.32309.8916428.2NoneNoneNoneNone
1267None1333384138437090NoneNone...3323.35170141364.517390.5-0.5-0.5NoneNoneNoneNone
\n", "

8 rows × 22 columns

\n", "
" ], "text/plain": [ " geom point_id source_measure_id destin_measure_id source source_idx \\\n", "id \n", "1178 None 788 2324 2325 7 0 \n", "1228 None 1269 3678 3677 7 0 \n", "1230 None 1271 3684 3683 7 0 \n", "1214 None 1243 3610 3609 7 0 \n", "1222 None 1261 3653 3652 7 0 \n", "1223 None 1262 3655 3654 7 0 \n", "1264 None 1328 3829 3828 7 0 \n", "1267 None 1333 3841 3843 7 0 \n", "\n", " destination destination_idx lat lon ... source_apriori_x \\\n", "id ... \n", "1178 9 0 None None ... 2231.12 \n", "1228 9 0 None None ... 3082.24 \n", "1230 9 0 None None ... 3908.22 \n", "1214 9 0 None None ... 3983.41 \n", "1222 9 0 None None ... 2674.81 \n", "1223 9 0 None None ... 3078.79 \n", "1264 9 0 None None ... 4237.17 \n", "1267 9 0 None None ... 3323.35 \n", "\n", " source_apriori_y destination_x destination_y destination_apriori_x \\\n", "id \n", "1178 16213.7 180.316 16595.5 204.205 \n", "1228 6807.21 1001.12 7160.22 978.351 \n", "1230 6813.23 1773.38 7180.67 1797.61 \n", "1214 1699.58 1877.61 2095.96 1840.91 \n", "1222 5831.07 493.5 6213.5 -0.5 \n", "1223 5826.54 893.047 6201.25 918.537 \n", "1264 16023.8 2321.67 16425.3 2309.89 \n", "1267 17014 1364.5 17390.5 -0.5 \n", "\n", " destination_apriori_y shift_x shift_y original_destination_x \\\n", "id \n", "1178 16599.5 None None None \n", "1228 7163.4 None None None \n", "1230 7184.37 None None None \n", "1214 2099.48 None None None \n", "1222 -0.5 None None None \n", "1223 6205.46 None None None \n", "1264 16428.2 None None None \n", "1267 -0.5 None None None \n", "\n", " original_destination_y \n", "id \n", "1178 None \n", "1228 None \n", "1230 None \n", "1214 None \n", "1222 None \n", "1223 None \n", "1264 None \n", "1267 None \n", "\n", "[8 rows x 22 columns]" ] }, "execution_count": 19, "metadata": {}, "output_type": "execute_result" } ], "source": [ "failed_both = edge_9_7.matches.loc[(~edge_9_7.masks[['fundamental', 'homography']]).all(axis=1)]\n", "failed_both" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "### Viewing individual matches\n", "Let's take a closer look at the matches that failed both checks. We're going to use the Roi (Region of Interest) objects in Autocnet that allow you to look at a small portion of an image." ] }, { "cell_type": "code", "execution_count": 20, "metadata": { "scrolled": false }, "outputs": [ { "data": { "image/png": "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\n", "text/plain": [ "
" ] }, "metadata": { "needs_background": "light" }, "output_type": "display_data" }, { "data": { "image/png": "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\n", "text/plain": [ "
" ] }, "metadata": { "needs_background": "light" }, "output_type": "display_data" }, { "data": { "image/png": "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\n", "text/plain": [ "
" ] }, "metadata": { "needs_background": "light" }, "output_type": "display_data" }, { "data": { "image/png": "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\n", "text/plain": [ "
" ] }, "metadata": { "needs_background": "light" }, "output_type": "display_data" }, { "data": { "image/png": "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\n", "text/plain": [ "
" ] }, "metadata": { "needs_background": "light" }, "output_type": "display_data" }, { "data": { "image/png": "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\n", "text/plain": [ "
" ] }, "metadata": { "needs_background": "light" }, "output_type": "display_data" }, { "data": { "image/png": "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\n", "text/plain": [ "
" ] }, "metadata": { "needs_background": "light" }, "output_type": "display_data" }, { "data": { "image/png": "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\n", "text/plain": [ "
" ] }, "metadata": { "needs_background": "light" }, "output_type": "display_data" } ], "source": [ "roi_size = 25\n", "for idx, match in failed_both.iterrows():\n", " with ncg.session_scope() as session:\n", " point_name = session.query(Points).filter(Points.id == match[\"point_id\"]).first().identifier\n", "\n", " source_image = ncg.nodes[match['source']]['data']\n", " dest_image = ncg.nodes[match['destination']]['data']\n", " source_roi = Roi(source_image.geodata, match['source_x'], match['source_y'], size_x=roi_size, size_y=roi_size)\n", " dest_roi = Roi(dest_image.geodata, match['destination_x'], match['destination_y'], size_x=roi_size, size_y=roi_size)\n", "\n", " fig, (ax1, ax2) = plt.subplots(1,2)\n", " fig.suptitle(point_name)\n", " ax1.imshow(source_roi.array, cmap='gray')\n", " ax1.plot(source_roi.center[0] + source_roi.axr, source_roi.center[1] + source_roi.ayr, 'ro')\n", " ax1.title.set_text(os.path.split(source_image['image_name'])[-1])\n", " ax2.imshow(dest_roi.array, cmap='gray')\n", " ax2.plot(dest_roi.center[0] + dest_roi.axr, dest_roi.center[1] + dest_roi.ayr, 'bo')\n", " ax2.title.set_text(os.path.split(dest_image['image_name'])[-1])\n", " fig.show()" ] }, { "cell_type": "code", "execution_count": null, "metadata": { "collapsed": true }, "outputs": [], "source": [] } ], "metadata": { "kernelspec": { "display_name": "Python 3", "language": "python", "name": "python3" }, "language_info": { "codemirror_mode": { "name": "ipython", "version": 3 }, "file_extension": ".py", "mimetype": "text/x-python", "name": "python", "nbconvert_exporter": "python", "pygments_lexer": "ipython3", "version": "3.8.3" } }, "nbformat": 4, "nbformat_minor": 4 }