def test_crop_panels():
"""
Test the cropPanels() function.
Returns
-------
None.
"""
#Pick the file name to read
img_file = "./examples/Panel_Detection_Examples/sat_img.png"
#Create an instance of the PanelDetection() class.
pc = pan_det.PanelDetection(model_file_path = './panel_segmentation/VGG16Net_ConvTranpose_complete.h5',
classifier_file_path = './panel_segmentation/VGG16_classification_model.h5')
#Read in the image
x = imagex.load_img(img_file,
color_mode='rgb',
target_size=(640,640))
x = np.array(x)
#Mask the satellite image
res = pc.testSingle(x.astype(float), test_mask=None, model =None)
#Crop the panels
new_res = pc.cropPanels(x, res)
#Assert that the 'new_res' variable is a numpy array and the dimensions.
assert (type(new_res) == np.ndarray) & (new_res.shape == (1, 640, 640, 3))
def test_cluster_panels():
"""
Test the clusterPanels() function.
Returns
-------
None.
"""
#Pick the file name to read
img_file = "./examples/Panel_Detection_Examples/sat_img.png"
#Create an instance of the PanelDetection() class.
pc = pan_det.PanelDetection(model_file_path = './panel_segmentation/VGG16Net_ConvTranpose_complete.h5',
classifier_file_path = './panel_segmentation/VGG16_classification_model.h5')
#Read in the image
x = imagex.load_img(img_file,
color_mode='rgb',
target_size=(640,640))
x = np.array(x)
#Mask the satellite image
res = pc.testSingle(x.astype(float), test_mask=None, model =None)
#Use the mask to isolate the panels
new_res = pc.cropPanels(x, res)
number_arrays = 5
clusters = pc.clusterPanels(new_res, res,
number_arrays)
azimuth_list = []
for ii in np.arange(clusters.shape[0]):
az = pc.detectAzimuth(clusters[ii][np.newaxis,:])
azimuth_list.append(az)
assert all(isinstance(x, int) for x in azimuth_list)
def test_plot_az():
"""
Test the detectAzimuth() function.
Returns
-------
None.
"""
#Pick the file name to read
img_file = "./examples/Panel_Detection_Examples/sat_img.png"
#Create an instance of the PanelDetection() class.
pc = pan_det.PanelDetection(model_file_path = './panel_segmentation/VGG16Net_ConvTranpose_complete.h5',
classifier_file_path = './panel_segmentation/VGG16_classification_model.h5')
#Read in the image
x = imagex.load_img(img_file,
color_mode='rgb',
target_size=(640,640))
x = np.array(x)
#Mask the satellite image
res = pc.testSingle(x.astype(float), test_mask=None, model =None)
#Crop the panels
new_res = pc.cropPanels(x, res)
pc.plotEdgeAz(new_res, 10, 1,
save_img_file_path = "./tests/",
plot_show = True)
#Open the image and assert that it exists
im = PIL.Image.open("./tests/crop_mask_az_0.png")
assert (type(im) == PIL.PngImagePlugin.PngImageFile)
def test_estimate_az():
"""
Test the detectAzimuth() function.
Returns
-------
None.
"""
#Pick the file name to read
img_file = "./examples/Panel_Detection_Examples/sat_img.png"
#Create an instance of the PanelDetection() class.
pc = pan_det.PanelDetection(model_file_path = './panel_segmentation/VGG16Net_ConvTranpose_complete.h5',
classifier_file_path = './panel_segmentation/VGG16_classification_model.h5')
#Read in the image
x = imagex.load_img(img_file,
color_mode='rgb',
target_size=(640,640))
x = np.array(x)
#Mask the satellite image
res = pc.testSingle(x.astype(float), test_mask=None, model =None)
#Crop the panels
new_res = pc.cropPanels(x, res)
az = pc.detectAzimuth(new_res)
#Assert that the azimut returned is an int instance
assert_isinstance(az, int)
def panelDetectionClass():
'''Generate an instance of the PanelDetection() class to run unit
tests on.'''
# Create an instance of the PanelDetection() class.
pc = pan_det.PanelDetection(
'./panel_segmentation/VGG16Net_ConvTranpose_complete.h5',
'./panel_segmentation/VGG16_classification_model.h5',
'./panel_segmentation/object_detection_model.pth')
return pc
def test_generate_satellite_image():
"""
Test the generateSatelliteImage() function.
Returns
-------
None.
"""
with pytest.raises(ValueError):
#Variables
latitude = 39.7407
longitude = -105.1694
google_maps_api_key = "Wrong_API_key"
file_name_save = "./examples/Panel_Detection_Examples/sat_img.png"
#Create an instance of the PanelDetection() class.
pc = pan_det.PanelDetection(model_file_path = './panel_segmentation/VGG16Net_ConvTranpose_complete.h5',
classifier_file_path = './panel_segmentation/VGG16_classification_model.h5')
pc.generateSatelliteImage(latitude, longitude,
file_name_save, google_maps_api_key)
def test_has_panels():
"""
Test the hasPanels() function.
Returns
-------
Assert that the value returned is a boolean
"""
#Pick the file name to read
img_file = "./examples/Panel_Detection_Examples/sat_img.png"
#Create an instance of the PanelDetection() class.
pc = pan_det.PanelDetection(model_file_path = './panel_segmentation/VGG16Net_ConvTranpose_complete.h5',
classifier_file_path = './panel_segmentation/VGG16_classification_model.h5')
#Read in the image
x = imagex.load_img(img_file,
color_mode='rgb',
target_size=(640,640))
x = np.array(x)
#Assert that the returned value is a boolean
panel_loc = pc.hasPanels(x)
assert_isinstance(panel_loc, bool)
def testGenerateSatelliteImage():
latitude = 39.7407
longitude = -105.1694
google_maps_api_key = "Wrong_API_key"
file_name_save = "./examples/Panel_Detection_Examples/sat_img.png"
model_file_path = os.path.abspath(
'./panel_segmentation/VGG16Net_ConvTranpose_complete.h5')
assert os.path.exists(model_file_path)
classifier_file_path = os.path.abspath(
'./panel_segmentation/VGG16_classification_model.h5')
assert os.path.exists(classifier_file_path)
mounting_classifier_model_path = os.path.abspath(
'./panel_segmentation/object_detection_model.pth')
assert os.path.exists(mounting_classifier_model_path)
# Create an instance of the PanelDetection() class.
pc = pan_det.PanelDetection(
model_file_path=model_file_path,
classifier_file_path=classifier_file_path,
mounting_classifier_file_path=mounting_classifier_model_path)
with pytest.raises(ValueError):
pc.generateSatelliteImage(latitude, longitude,
file_name_save, google_maps_api_key)