Test/intrinsics input regressions

main_calcIntrinsics.py

@@ -26,6 +26,9 @@
 saveIntrinsicsForDeployment = True
 
 deployedFolderNames = ['Deployed_720_60fps','Deployed'] # both folder names if want to keep the detailed folder
+
+cameraModel = "iPhoneTest"  
+videoType = ".mov" #can be .avi or other file formats
 
 # %% Paths to data folder for local testing.
 dataDir = os.path.join(getDataDirectory(),'Data')
@@ -34,7 +37,6 @@
 intrinsicComparisonFile = os.path.join(sessionDir,'intrinsicComparison.pkl')
 
 # %% Get checker parameters and filenames if they exist
-
 # TODO this should come from the server API
 # Get checkerboard parameters from metadata.
 metadataPath = os.path.join(sessionDir,'sessionMetadata.yaml')
@@ -61,7 +63,7 @@
 
 
 # Compute average intrinsic values from multiple trials of same camera
-CamParamsAverage, CamParamList, intrinsicComparisons, cameraModel = computeAverageIntrinsics(sessionDir,trials,CheckerBoardParams,nImages=50)
+CamParamsAverage, CamParamList, intrinsicComparisons, cameraModel = computeAverageIntrinsics(sessionDir,trials,CheckerBoardParams,nImages=50,cameraModel=cameraModel,videoType=videoType)
 
 
 # Save intrinsics from first camera for deployement 

tests/test_calibration.py

@@ -1,6 +1,8 @@
 import os
 import shutil
 import sys
+from pathlib import Path
+from unittest.mock import Mock
 
 import cv2
 import numpy as np
@@ -48,6 +50,10 @@
     'production_success',
     'utah_production_success.mov',
 )
+IPHONE_17E_FIXTURE_DIR = os.path.join(
+    REPO_DIR,
+    'a8d94635-77db-4be4-ab7f-25da84c1a092',
+)
 
 PRIMARY_SUCCESS_FIXTURES = [
     (
@@ -418,3 +424,134 @@ def test_negative_fixtures_reject(
     assert calls['fallback'] > 0
     assert fallback_flags
     assert all(flags & cv2.CALIB_CB_EXHAUSTIVE for flags in fallback_flags)
+
+
+# ---- Intrinsics workflow regression tests ----
+
+def make_intrinsics(focal_length):
+    return {
+        'intrinsicMat': np.array([
+            [focal_length, 0.0, 320.0],
+            [0.0, focal_length + 10.0, 240.0],
+            [0.0, 0.0, 1.0],
+        ]),
+        'distortion': np.array([[0.1, -0.1, 0.01, 0.02, 0.03]]),
+        'imageSize': np.array([[480.0], [640.0]]),
+    }
+
+
+def test_intrinsics_api(tmp_path, monkeypatch):
+    video_url = 'https://example.test/trial-a.mov'
+    response = Mock()
+    response.json.return_value = {
+        'name': 'null',
+        'videos': [{
+            'video': video_url,
+            'parameters': {'model': 'iPhone13,3'},
+        }],
+    }
+    request = Mock(return_value=response)
+    download = Mock(side_effect=lambda _, path: Path(path).touch())
+    extraction = Mock()
+    params = make_intrinsics(1000.0)
+    monkeypatch.setattr(utilsChecker, 'makeRequestWithRetry', request)
+    monkeypatch.setattr(utilsChecker, 'download_file', download)
+    monkeypatch.setattr(utilsChecker, 'video2Images', extraction)
+    monkeypatch.setattr(utilsChecker, 'calcIntrinsics', Mock(return_value=params))
+
+    average, _, _, model = utilsChecker.computeAverageIntrinsics(
+        str(tmp_path), ['trial-a'], DEFAULT_CHECKERBOARD_PARAMS
+    )
+
+    expected_path = os.path.join(tmp_path, 'trial-a', 'trial-a.mov')
+    request.assert_called_once()
+    download.assert_called_once_with(video_url, expected_path)
+    assert extraction.call_args.args[0] == expected_path
+    assert model == 'iPhone13,3'
+    np.testing.assert_allclose(average['intrinsicMat'], params['intrinsicMat'])
+
+
+def test_intrinsics_local(tmp_path, monkeypatch):
+    trial_ids = ['capture-a', 'capture-b']
+    paths = []
+    for name in trial_ids:
+        path = tmp_path / name / f'{name}.avi'
+        path.parent.mkdir()
+        path.touch()
+        paths.append(str(path))
+
+    request = Mock()
+    download = Mock()
+    extraction = Mock()
+    monkeypatch.setattr(utilsChecker, 'makeRequestWithRetry', request)
+    monkeypatch.setattr(utilsChecker, 'download_file', download)
+    monkeypatch.setattr(utilsChecker, 'video2Images', extraction)
+    monkeypatch.setattr(
+        utilsChecker,
+        'calcIntrinsics',
+        Mock(side_effect=[make_intrinsics(900.0), make_intrinsics(1100.0)]),
+    )
+
+    average, _, _, model = utilsChecker.computeAverageIntrinsics(
+        str(tmp_path),
+        trial_ids,
+        DEFAULT_CHECKERBOARD_PARAMS,
+        nImages=5,
+        cameraModel='ResearchCamera',
+        videoType='.avi',
+    )
+
+    request.assert_not_called()
+    download.assert_not_called()
+    assert [call.args[0] for call in extraction.call_args_list] == paths
+    assert model == 'ResearchCamera'
+    np.testing.assert_allclose(average['intrinsicMat'][0, 0], 1000.0)
+
+
+# This fixture is shared separately and is not tracked in this repository, so
+# this regression runs locally when available and skips elsewhere (including CI).
+@pytest.mark.skipif(
+    not os.path.isdir(IPHONE_17E_FIXTURE_DIR),
+    reason='iPhone 17e calibration fixture is not available',
+)
+def test_intrinsics_iphone17e():
+    average, captures, _, model = utilsChecker.computeAverageIntrinsics(
+        IPHONE_17E_FIXTURE_DIR,
+        ['iphone17e', 'iphone17e_1', 'iphone17e_2'],
+        {'dimensions': (11, 8), 'squareSize': 60},
+        cameraModel='iPhone18,5',
+        videoType='.mov',
+    )
+    expected = loadCameraParameters(os.path.join(
+        REPO_DIR,
+        'CameraIntrinsics',
+        'iPhone18,5',
+        'Deployed',
+        'cameraIntrinsics.pickle',
+    ))
+
+    assert len(captures) == 3
+    assert model == 'iPhone18,5'
+    for name in expected:
+        np.testing.assert_allclose(average[name], expected[name])
+
+
+# PR #281 currently raises UnboundLocalError here. Keep this skipped test as
+# possible future coverage for clearer missing-local-video handling.
+@pytest.mark.skip(reason='missing local videos are not handled by PR #281')
+def test_intrinsics_missing(tmp_path, monkeypatch):
+    request = Mock()
+    download = Mock()
+    monkeypatch.setattr(utilsChecker, 'makeRequestWithRetry', request)
+    monkeypatch.setattr(utilsChecker, 'download_file', download)
+
+    with pytest.raises(FileNotFoundError, match='capture-a\\.avi'):
+        utilsChecker.computeAverageIntrinsics(
+            str(tmp_path),
+            ['capture-a'],
+            DEFAULT_CHECKERBOARD_PARAMS,
+            cameraModel='ResearchCamera',
+            videoType='.avi',
+        )
+    request.assert_not_called()
+    download.assert_not_called()

utilsChecker.py

@@ -194,24 +194,29 @@ def calcIntrinsics(folderName, CheckerBoardParams=None, filenames=['*.jpg'],
     return CamParams
 
 # %%
-def computeAverageIntrinsics(session_path,trialIDs,CheckerBoardParams,nImages=25):
+def computeAverageIntrinsics(session_path,trialIDs,CheckerBoardParams,nImages=25,cameraModel= None,videoType=".mov"):
     CamParamList = []
     camModels = []
+    trial_name = 'null'
 
     for trial_id in trialIDs:
-        resp = makeRequestWithRetry('GET',
-                                    API_URL + "trials/{}/".format(trial_id),
-                                    headers = {"Authorization": "Token {}".format(API_TOKEN)})
-        trial = resp.json()
-        camModels.append(trial['videos'][0]['parameters']['model'])
-        trial_name = trial['name']
+        if cameraModel is None:
+            resp = makeRequestWithRetry(
+                'GET', API_URL + "trials/{}/".format(trial_id),
+                headers={"Authorization": "Token {}".format(API_TOKEN)})
+            trial = resp.json()
+            camModels.append(trial['videos'][0]['parameters']['model'])
+            trial_name = trial['name']
+        else:
+            camModels.append(cameraModel)
+            trial_name = trial_id
         if trial_name == 'null':
             trial_name = trial_id
 
         # Make directory (folder for trialname, intrinsics also saved there)
         video_dir = os.path.join(session_path,trial_name)
         os.makedirs(video_dir, exist_ok=True)
-        video_path = os.path.join(video_dir,trial_name + ".mov")
+        video_path = os.path.join(video_dir,trial_name + videoType)
 
         # Download video if not done
         if not os.path.exists(video_path):