def get_visualization(self, analysis, result_path, data_path, project,
main_window):
"""
This function should show the complete Visualization
"""
w = QWidget()
w.setLayout(QVBoxLayout(w))
w.layout().addWidget(
QLabel(
"Color CIE-Lab:".rjust(20) +
str(analysis.get_adata()['color_lab']), w))
w.layout().addWidget(
QLabel(
"Color BGR:".rjust(20) +
str(analysis.get_adata()['color_bgr']), w))
w.layout().addWidget(
QLabel(
"Saturation Luebbe:".rjust(20) +
str(analysis.get_adata()['saturation_l']), w))
w.layout().addWidget(
QLabel(
"Saturation FilmCo:".rjust(20) +
str(analysis.get_adata()['saturation_p']), w))
view = EGraphicsView(w)
view.set_image(
numpy_to_pixmap(
np.array(([[analysis.get_adata()['color_bgr']] * 100] *
25)).astype(np.uint8)))
w.layout().addWidget(view)
return [
VisualizationTab(widget=w,
name="Color-Features",
use_filter=False,
controls=None)
]
def set_img_movie(self, img):
self.img_movie = img
if img is None:
return
if self.project is not None and self.project.headless_mode:
return
fx = CACHE_WIDTH / img.shape[1]
img = cv2.resize(img, None, None, fx, fx, cv2.INTER_CUBIC)
self.img_movie = img
if img.shape[2] == 3:
self.onImageSet.emit(self, self.img_movie, numpy_to_pixmap(img))
elif img.shape[2] == 4:
self.onImageSet.emit(
self, self.img_movie,
numpy_to_pixmap(img, cvt=cv2.COLOR_BGRA2RGBA, with_alpha=True))
def get_preview(self, analysis: IAnalysisJobAnalysis):
"""
This should return the Widget that is shown in the Inspector when the analysis is selected
"""
w = QWidget()
w.setLayout(QVBoxLayout(w))
w.layout().addWidget(
QLabel(
"Color CIE-Lab:".rjust(20) +
str(analysis.get_adata()['color_lab']), w))
w.layout().addWidget(
QLabel(
"Color BGR:".rjust(20) +
str(analysis.get_adata()['color_bgr']), w))
w.layout().addWidget(
QLabel(
"Saturation Luebbe:".rjust(20) +
str(analysis.get_adata()['saturation_l']), w))
w.layout().addWidget(
QLabel(
"Saturation FilmCo:".rjust(20) +
str(analysis.get_adata()['saturation_p']), w))
view = EGraphicsView(w)
view.set_image(
numpy_to_pixmap(
np.array(([[analysis.get_adata()['color_bgr']] * 100] *
25)).astype(np.uint8)))
w.layout().addWidget(view)
return w
def get_visualization(self, analysis, result_path, data_path, project,
main_window):
"""
This function should show the complete Visualization
"""
widget = EGraphicsView(None, auto_frame=True, has_context_menu=True)
n = 20
colormap = get_colormap(n)
data = analysis.get_adata()
img = np.zeros(shape=data.shape + (3, ), dtype=np.float32)
for i in range(n):
img[data == i] = colormap[i][:3]
img = (img * 255).astype(np.uint8)
img = cv2.cvtColor(img, cv2.COLOR_RGBA2BGR)
if isinstance(analysis.target_container, Screenshot):
scr = analysis.target_container.get_img_movie_orig_size()
scr = cv2.resize(scr,
img.shape[:2][::-1],
interpolation=cv2.INTER_CUBIC).astype(np.float32)
img = ((scr * 0.7) + (img * 0.3)).astype(np.uint8)
widget.set_image(numpy_to_pixmap(img))
return [
VisualizationTab(widget=widget,
name="Semantic Segmentation Mask",
use_filter=False,
controls=None)
]
def get_preview(self, analysis: IAnalysisJobAnalysis):
"""
This should return the Widget that is shown in the Inspector when the analysis is selected
"""
widget = EGraphicsView(None, auto_frame=True)
n = 20
colormap = get_colormap(n)
data = analysis.get_adata()
img = np.zeros(shape=data.shape + (3, ), dtype=np.float32)
for i in range(n):
img[data == i] = colormap[i][:3]
img = (img * 255).astype(np.uint8)
img = cv2.cvtColor(img, cv2.COLOR_RGBA2BGR)
if isinstance(analysis.target_container, Screenshot):
scr = analysis.target_container.get_img_movie_orig_size()
scr = cv2.resize(scr,
img.shape[:2][::-1],
interpolation=cv2.INTER_CUBIC).astype(np.float32)
img = ((scr * 0.7) + (img * 0.3)).astype(np.uint8)
widget.set_image(numpy_to_pixmap(img))
return widget
def on_slider_change(self):
if self.cap is None:
return
self.cap.set(cv2.CAP_PROP_POS_FRAMES, int(self.pos_slider.value() * (self.duration / 1000)))
ret, frame = self.cap.read()
if frame is not None:
self.onFrameChanged.emit(numpy_to_pixmap(frame))
self.get_rect()
def get_preview(self, analysis: IAnalysisJobAnalysis):
"""
This should return the Widget that is shown in the Inspector when the analysis is selected
"""
view = EGraphicsView(None)
pixmap = numpy_to_pixmap(analysis.get_adata())
view.set_image(pixmap)
return view
def get_visualization(self, analysis, result_path, data_path, project, main_window):
"""
This function should show the complete Visualization
"""
view = EGraphicsView(None)
pixmap = numpy_to_pixmap(analysis.get_adata())
view.set_image(pixmap)
return [VisualizationTab(widget=view, name="Z-Projection", use_filter=False, controls=None)]
def get_visualization(self, analysis, result_path, data_path, project,
main_window):
"""
This function should show the complete Visualization
"""
widget = EGraphicsView(None, auto_frame=True)
widget.set_image(
numpy_to_pixmap(
self.barcode_to_image(analysis.get_adata()['barcode'])))
return [
VisualizationTab(widget=widget,
name="Barcode",
use_filter=False,
controls=None)
]
def get_preview(self, analysis: IAnalysisJobAnalysis):
"""
This should return the Widget that is shown in the Inspector when the analysis is selected
"""
if analysis.parameters['interpolation'] == "Cubic":
interpolation = cv2.INTER_CUBIC
else:
interpolation = cv2.INTER_LINEAR
image = self.barcode_to_image(analysis.get_adata()['barcode'],
400,
image_height=50,
interpolation=interpolation)
barcode_pixm = numpy_to_pixmap(image)
view = QGraphicsView(QGraphicsScene())
view.scene().addPixmap(barcode_pixm)
return view
def set_classification_object(self,
clobj,
recompute=False,
hdf5_cache=None):
if not recompute and clobj.unique_id in self.masked_cache:
result = self.masked_cache[clobj.unique_id]
elif clobj is None or clobj.semantic_segmentation_labels[0] == "":
result = self.img_movie
else:
result = None
cached = hdf5_cache.get_screenshot(clobj.get_id(), self.unique_id)
if cached is None:
a = self.get_semantic_segmentations(
clobj.semantic_segmentation_labels[0])
lbls = clobj.semantic_segmentation_labels[1]
if len(clobj.semantic_segmentation_labels[0]) == len(lbls):
result = self.img_movie
if a is not None:
mask = a.get_adata()
masked = apply_mask(self.img_movie, mask, lbls)
h = CACHE_WIDTH / masked.shape[0] * masked.shape[1]
masked = cv2.resize(masked, (int(h), CACHE_WIDTH),
interpolation=cv2.INTER_CUBIC)
hdf5_cache.dump_screenshot(clobj.get_id(), self.unique_id,
masked)
result = masked
else:
result = cached
if result is not None:
self.masked_cache[clobj.unique_id] = result
self.onImageSet.emit(
self, result,
numpy_to_pixmap(result,
cvt=cv2.COLOR_BGRA2RGBA,
with_alpha=True))
return result