pyscript-bootstrap-templates/examples/04_cell_detector/main.py

309 lines
11 KiB
Python
Raw Permalink Normal View History

2023-06-03 12:21:28 +02:00
from pyscript_bootstrap_templates import bootstrap_templates
from pyscript_bootstrap_templates import bootstrap_HTML as bHTML
from pyscript_bootstrap_templates import bootstrap_inputs as bInputs
from pyscript_bootstrap_templates import HTML as HTML
import numpy as np
from PIL import Image
import cv2
2023-06-04 18:51:06 +02:00
loaded_img:np.ndarray = None
2023-06-03 12:21:28 +02:00
2023-06-04 18:51:06 +02:00
app = bootstrap_templates.PyScriptBootstrapDashboard(parent_element="pyscript_app", brand_name="Pyscript Cell Colony Detector")
2023-06-03 12:21:28 +02:00
main_div = bHTML.BootstrapContainer(parent=app.main)
main_div.w = 100
2023-06-04 18:51:06 +02:00
result_div = bHTML.BootstrapContainer (parent=main_div)
2023-06-03 12:21:28 +02:00
def process_image(image,
hough_min_dist = 500,
hough_param1 = 80,
hough_param2 = 500,
minRadius = 100,
maxRadius=500,
2023-06-04 18:51:06 +02:00
inner_hough_param1 = 25,
inner_hough_param2 = 50,
inner_hough_circles=True,
cell_colony_color_channel:int = None):
2023-06-03 12:21:28 +02:00
for child in result_div.children:
child.destroy()
2023-06-04 18:51:06 +02:00
2023-06-03 12:21:28 +02:00
# Convert the image to grayscale
2023-06-04 18:51:06 +02:00
if cell_colony_color_channel is None:
gray = cv2.cvtColor(image, cv2.COLOR_BGR2GRAY)
else:
gray = image.astype(float)[...,cell_colony_color_channel]
gray = (255 * gray / gray.max()).astype(np.uint8)
2023-06-03 12:21:28 +02:00
# Use median blur to reduce noise
gray = cv2.medianBlur(gray, 5)
# Apply Hough transform on the image to find circles
circles = cv2.HoughCircles(gray,
cv2.HOUGH_GRADIENT,
2,
minDist=hough_min_dist,
param1=hough_param1,
param2=hough_param2,
minRadius=minRadius,
maxRadius=maxRadius)
if circles is None:
2023-06-04 18:51:06 +02:00
app.alert_danger("No cell areas detected in image")
2023-06-03 12:21:28 +02:00
return False
# Convert to integers
circles = np.uint16(np.around(circles))
# Loop over all detected circles and add them to the image
#for i in circles[0,:]:
# # Draw outer circle
# cv2.circle(image,(i[0],i[1]),i[2],(0,255,0),2)
# # Draw center of circle
# cv2.circle(image,(i[0],i[1]),2,(0,0,255),3)
first_wells = []
first_wells_gray = []
for i in circles[0,:]:
center = (i[0], i[1])
radius = i[2]
# Cut out the well
well = image[center[1]-radius:center[1]+radius, center[0]-radius:center[0]+radius]
well_gray = gray[center[1]-radius:center[1]+radius, center[0]-radius:center[0]+radius]
first_wells.append(well)
first_wells_gray.append(well_gray)
if inner_hough_circles:
second_wells = []
second_wells_gray = []
for well, well_gray in zip(first_wells, first_wells_gray):
2023-06-04 18:51:06 +02:00
circles = cv2.HoughCircles(well_gray,
cv2.HOUGH_GRADIENT,
1,
minDist=hough_min_dist,
param1=inner_hough_param1,
param2=inner_hough_param2,
minRadius=int((min(well.shape[0],well.shape[1]) // 2)*0.79),
maxRadius=int((min(well.shape[0],well.shape[1]) // 2)*0.95))
2023-06-03 12:21:28 +02:00
if circles is not None:
circles = np.uint16(np.around(circles))
i = circles[0,:][0]
center = (i[0], i[1])
radius = i[2] - 1
min_y = max(center[1]-radius, 0)
max_y = center[1]+radius
min_x = max(center[0]-radius, 0)
max_x = center[0]+radius
second_wells.append(well[min_y:max_y, min_x:max_x])
second_wells_gray.append(well_gray[min_y:max_y, min_x:max_x])
else:
second_wells.append(well)
second_wells_gray.append(well_gray)
else:
second_wells = first_wells
second_wells_gray = first_wells_gray
2023-06-04 18:51:06 +02:00
tabs = {}
2023-06-03 12:21:28 +02:00
2023-06-04 18:51:06 +02:00
for well in second_wells:
2023-06-03 12:21:28 +02:00
2023-06-04 18:51:06 +02:00
well_gray = cv2.cvtColor(well, cv2.COLOR_BGR2GRAY)
2023-06-03 12:21:28 +02:00
_, binary = cv2.threshold(255 - well_gray, 0, 255, cv2.THRESH_BINARY + cv2.THRESH_OTSU)
# Perform some morphological operations to remove small noise - you can change the kernel size
kernel = np.ones((3,3),np.uint8)
cleaned = cv2.morphologyEx(binary, cv2.MORPH_OPEN, kernel)
2023-06-04 18:51:06 +02:00
radius = min(well.shape[0], well.shape[1]) // 2 -1
2023-06-03 12:21:28 +02:00
# Now clear all pixels outside the circular well.
# We do this by creating a mask for the well and applying it to the image.
well_mask = np.zeros_like(cleaned, dtype=np.uint8)
2023-06-04 18:51:06 +02:00
cv2.circle(well_mask, (well.shape[1] // 2, well.shape[0] // 2), radius, 1, thickness=-1)
2023-06-03 12:21:28 +02:00
cleaned = cleaned * well_mask
circle = well_mask > 0
# now create an image overlay to display
ratio = np.sum((cleaned > 0).astype(int)) / np.sum(circle.astype(int))
2023-06-04 18:51:06 +02:00
div_result = bHTML.BootstrapContainer(f"ratio: {ratio * 100}%")
2023-06-03 12:21:28 +02:00
div_result.shadow = bHTML.Shadow.LARGE
div_result.rounded = True
2023-06-04 18:51:06 +02:00
div_result.w = 75
div_result.h = 75
div_result.rounded_size = 50
div_result.p = 3
2023-06-03 12:21:28 +02:00
2023-06-04 18:51:06 +02:00
well = well.copy()
2023-06-03 12:21:28 +02:00
well[cleaned > 0,0] = 255
final_img = bHTML.Image.from_numpy_array(well, parent=div_result)
final_img.rounded = True
2023-06-04 18:51:06 +02:00
final_img.rounded_size = 50
2023-06-03 12:21:28 +02:00
final_img.shadow = bHTML.Shadow.LARGE
final_img.w = 100
2023-06-04 18:51:06 +02:00
tabs[f"Well #{len(tabs) + 1}"] = div_result
tabs = bHTML.Tabs(tabs, parent=result_div)
tabs.w = 100
2023-06-03 12:21:28 +02:00
div = bHTML.BootstrapContainer("Controls", parent=app.sidebar)
div.font_size = 4
image_input = bInputs.InputFile(label_text="choose image file", parent=app.sidebar)
2023-06-04 18:51:06 +02:00
btn = bHTML.ButtonPrimary("Process", parent=app.sidebar)
btn.w = 100
btn.mt = 3
btn.mb = 3
btn.ml = 4
btn.mr = 4
i_hough_min_dist = bInputs.InputInt("hough min distance [px]",
parent=app.sidebar,
help_text="minimal distance between wells in pixels")
2023-06-03 12:21:28 +02:00
i_hough_min_dist.value = 500
2023-06-04 18:51:06 +02:00
i_hough_param1 = bInputs.InputInt("hough param1",
parent=app.sidebar,
help_text="parameter for canny edge detector")
2023-06-03 12:21:28 +02:00
i_hough_param1.value = 80
2023-06-04 18:51:06 +02:00
i_hough_param2 = bInputs.InputInt("hough param2",
parent=app.sidebar,
help_text="increase this value to prevent false circle detection")
2023-06-03 12:21:28 +02:00
i_hough_param2.value = 500
2023-06-04 18:51:06 +02:00
i_hough_min_radius = bInputs.InputInt("hough min radius [px]",
parent=app.sidebar,
help_text="min radius for circle detection")
2023-06-03 12:21:28 +02:00
i_hough_min_radius.value = 100
2023-06-04 18:51:06 +02:00
i_hough_max_radius = bInputs.InputInt("hough max radius [px]",
parent=app.sidebar,
help_text="max radius for circle detection")
2023-06-03 12:21:28 +02:00
i_hough_max_radius.value = 500
2023-06-04 18:51:06 +02:00
i_inner_hough = bInputs.InputCheckboxSingle("nested hough transform?",
parent=app.sidebar,
help_text="if set, circle detection will be applied twice on detected wells")
i_inner_hough.value = True
i_inner_hough.m = 3
i_inner_hough_param1 = bInputs.InputInt("inner hough param1", parent=app.sidebar)
i_inner_hough_param1.value = 25
i_inner_hough_param2 = bInputs.InputInt("inner hough param2", parent=app.sidebar)
i_inner_hough_param2.value = 50
i_use_color_channel = bInputs.InputSelect(["all",
"red",
"green",
"blue"],
label_text="cell colony color channel",
parent=app.sidebar,
help_text="if cells are more present in a specific channel, select it here")
i_use_color_channel.value = "blue"
2023-06-03 12:21:28 +02:00
def on_image_change(f, *args):
2023-06-04 18:51:06 +02:00
try:
global loaded_img
f.seek(0)
img = np.array(Image.open(f))
for child in result_div.children:
child.destroy()
bHTML.BootstrapContainer("input image:", parent=result_div)
output = bHTML.Image.from_numpy_array(img, parent=result_div)
output.rounded = True
output.rounded_size = 10
output.shadow = bHTML.Shadow.LARGE
output.width = "100%"
output.height = "100%"
loaded_img = img
app.toast("successfully loaded image", "info")
#toast = bHTML.Toast("successfully loaded image", title="info", parent=toast_container)
#toast.animation = True
#toast.show()
except Exception as e:
for child in result_div.children:
child.destroy()
bHTML.AlertDanger(f"error while loading image: {str(e)}", parent=alert_container)
2023-06-03 12:21:28 +02:00
image_input.onchange = on_image_change
def on_click(*args, **kwargs):
if loaded_img is None:
2023-06-04 18:51:06 +02:00
app.alert_danger("No image loaded")
2023-06-03 12:21:28 +02:00
return
2023-06-04 18:51:06 +02:00
for child in result_div.children:
child.destroy()
try:
h_min_dist = int(i_hough_min_dist.value)
h_param1 = int(i_hough_param1.value)
h_param2 = int(i_hough_param2.value)
h_inner_param1 = int(i_inner_hough_param1.value)
h_inner_param2 = int(i_inner_hough_param2.value)
h_min_radius = int(i_hough_min_radius.value)
h_max_radius = int(i_hough_max_radius.value)
inner_hough = bool(i_inner_hough.value)
color_channel = {
"all": None,
"red": 0,
"green": 1,
"blue": 2
}[i_use_color_channel.value]
process_image(loaded_img,
hough_min_dist=h_min_dist,
hough_param1=h_param1,
hough_param2=h_param2,
minRadius=h_min_radius,
maxRadius=h_max_radius,
inner_hough_param1=h_inner_param1,
inner_hough_param2=h_inner_param2,
inner_hough_circles=inner_hough,
cell_colony_color_channel=color_channel)
app.toast("successfully processed image", title="info")
except Exception as e:
for child in result_div.children:
child.destroy()
app.alert_danger(f"error while processing image: {str(e)}")
2023-06-03 12:21:28 +02:00
btn.onclick = on_click