Skip to content

1 improve bayesian optimizer code #2

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Draft
sandeep-ps wants to merge 8 commits into
base: main
Choose a base branch
from
Draft

1 improve bayesian optimizer code #2

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
8 commits
Select commit Hold shift + click to select a range
9da228b
add ML
bingzhang Feb 13, 2025
338bd41
add ML
bingzhang Feb 13, 2025
b488c17
send file id
bingzhang Feb 26, 2025
529f6c0
average runs
bingzhang Apr 21, 2025
9812ce3
update code
bingzhang Apr 23, 2025
a7a5926
integrate printability
bingzhang May 20, 2025
df16091
skip failed print
bingzhang Aug 1, 2025
c45c172
handle failed prints
bingzhang Aug 3, 2025
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
22 changes: 22 additions & 0 deletions Dockerfile
View file
Open in desktop
Original file line number Diff line number Diff line change
@@ -0,0 +1,22 @@
FROM python:3.9

RUN apt-get update && apt-get install -y \
libgl1-mesa-glx \
libglib2.0-0 \
libsm6 \
libxrender-dev \
libxext6 \
&& rm -rf /var/lib/apt/lists/*

WORKDIR /home/clowder

RUN python -m pip install --upgrade pip
#RUN python -m pip install pyclowder

COPY requirements.txt ./
#RUN pip install -r requirements.txt
RUN pip install --no-cache-dir -r requirements.txt

COPY printability.py optimizer.py image_analysis_extractor.py Jim_ColorHistogram_ColorScatterPlot.py AnalysisToolbox_Jim.py ColorHistogram_ColorScatterPlot.py extractor_info.json ./

CMD python3 image_analysis_extractor.py
327 changes: 166 additions & 161 deletions Jim_ColorHistogram_ColorScatterPlot.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -66,207 +66,212 @@ def getBiggestContour(contours, num_contours=1):
except:
print(traceback.format_exc())

def image_analysis(filename):
#--------------------------------------------------imageFile is some RGB image
#'Green_Sanghyun.jpg'
imageFile = os.path.join(filename)

#--------------------------------------------------imageFile is some RGB image

imageFile = os.path.join('Xiao_1um.png')

pix2 = cv.imread(imageFile)

#-------------------------------------------------Convert the BRG image to RGB
img = cv.cvtColor(pix2, cv.COLOR_BGR2RGB)

try:
# convert to grayscale before thresholding
#(we *could* do color thresholding as well if necessary)

grayscale_img = cv.cvtColor(img, cv.COLOR_RGB2GRAY)

# threshold *inverted* image (first number is cutoff value,
# second number assigned to all pixels exceeding cutoff)

bin_img = cv.threshold(grayscale_img,
25,
255,
cv.THRESH_BINARY_INV)[1]

# get list of contours:
cntrs = getContours(cv.bitwise_not(bin_img))

# keep biggest contour only:
cntr = getBiggestContour(cntrs)

# show contour:
img_cntr = img.copy()
cv.drawContours(img_cntr, cntr, -1, color=(255, 0, 0), thickness=3)
fig0 = plt.imshow(img_cntr)

# create blank mask:
mask = np.full(bin_img.shape, 0, dtype=np.uint8)

# use biggest contour to define ROI area:
cv.drawContours(mask, cntr, -1, (255, 255, 255), cv.FILLED)

#---------------------------------------------Convert the RGB image to HSV
pix2 = cv.cvtColor(img, cv.COLOR_RGB2HSV)
# --------------------------------------------------Splitting HSV channels
h, s, v = cv.split(pix2)

except:
print(traceback.format_exc())
pix2 = cv.imread(imageFile)

#--------------------------------Making some empty matrices of the same size
#-------------------------------------------------Convert the BRG image to RGB
img = cv.cvtColor(pix2, cv.COLOR_BGR2RGB)

pix_test = np.zeros((pix2.shape[0],pix2.shape[1]))
pix_mid = np.zeros((pix2.shape[0],pix2.shape[1]))
try:
# convert to grayscale before thresholding
#(we *could* do color thresholding as well if necessary)

grayscale_img = cv.cvtColor(img, cv.COLOR_RGB2GRAY)

# threshold *inverted* image (first number is cutoff value,
# second number assigned to all pixels exceeding cutoff)

bin_img = cv.threshold(grayscale_img,
25,
255,
cv.THRESH_BINARY_INV)[1]

# get list of contours:
cntrs = getContours(cv.bitwise_not(bin_img))

# keep biggest contour only:
cntr = getBiggestContour(cntrs)

# show contour:
img_cntr = img.copy()
cv.drawContours(img_cntr, cntr, -1, color=(255, 0, 0), thickness=3)
fig0 = plt.imshow(img_cntr)

# create blank mask:
mask = np.full(bin_img.shape, 0, dtype=np.uint8)

# use biggest contour to define ROI area:
cv.drawContours(mask, cntr, -1, (255, 255, 255), cv.FILLED)

#---------------------------------------------Convert the RGB image to HSV
pix2 = cv.cvtColor(img, cv.COLOR_RGB2HSV)
# --------------------------------------------------Splitting HSV channels
h, s, v = cv.split(pix2)

except:
print(traceback.format_exc())

#--------------------------------Making some empty matrices of the same size

pix_test = np.zeros((pix2.shape[0],pix2.shape[1]))
pix_mid = np.zeros((pix2.shape[0],pix2.shape[1]))

#----------------------------------------Creating a list of all the HSV values
flat_pix2=pix2.reshape((pix2.shape[1]*pix2.shape[0],3))
flat_mask = mask.reshape((mask.shape[1] * mask.shape[0]), 1)

#----------------------------------------Creating a list of all the RGB values
flat_img=img.reshape((img.shape[1]*img.shape[0],3))

#--------------------------------------------------------------------------
#----------------------------------------------------FILTERING-------------
#--------------------------------------------------------------------------
dictionary_HSV = {
'H':flat_pix2[:,0],
'S':flat_pix2[:,1],
'V':flat_pix2[:,2],
'filter':flat_mask[:,0]}

df = pd.DataFrame(dictionary_HSV)

#----------------------------------------Creating a list of all the HSV values
flat_pix2=pix2.reshape((pix2.shape[1]*pix2.shape[0],3))
flat_mask = mask.reshape((mask.shape[1] * mask.shape[0]), 1)

#----------------------------------------Creating a list of all the RGB values
flat_img=img.reshape((img.shape[1]*img.shape[0],3))
dictionary_RGB = {
'R':flat_img[:,0],
'G':flat_img[:,1],
'B':flat_img[:,2],
'filter':flat_mask[:,0]}

#--------------------------------------------------------------------------
#----------------------------------------------------FILTERING-------------
#--------------------------------------------------------------------------
dictionary_HSV = {
'H':flat_pix2[:,0],
'S':flat_pix2[:,1],
'V':flat_pix2[:,2],
'filter':flat_mask[:,0]}
dfimg = pd.DataFrame(dictionary_RGB)

df = pd.DataFrame(dictionary_HSV)

#------------------------------------Now filter just the image portion
#------------------------------------use "!=" instead of "==" to get background
df2 = df.loc[df['filter'] == 255]
df3 = dfimg.loc[dfimg['filter'] == 255]

dictionary_RGB = {
'R':flat_img[:,0],
'G':flat_img[:,1],
'B':flat_img[:,2],
'filter':flat_mask[:,0]}
#---------------------------------------------------------------------------
#-----------------------------------------------HISTOGRAM ------------------
#---------------------------------------------------------------------------
from scipy.stats import circmean, circstd

dfimg = pd.DataFrame(dictionary_RGB)
#-------------------------------------------------------------------------RGB
R_DIST = df3['R']
R_mu = np.mean(R_DIST)
R_sig = np.std(R_DIST)

G_DIST = df3['G']
G_mu = np.mean(G_DIST)
G_sig = np.std(G_DIST)

#------------------------------------Now filter just the image portion
#------------------------------------use "!=" instead of "==" to get background
df2 = df.loc[df['filter'] == 255]
df3 = dfimg.loc[dfimg['filter'] == 255]
B_DIST = df3['B']
B_mu = np.mean(B_DIST)
B_sig = np.std(B_DIST)

#---------------------------------------------------------------------------
#-----------------------------------------------HISTOGRAM ------------------
#---------------------------------------------------------------------------
from scipy.stats import circmean, circstd
fig1, axes = plt.subplots(1, 3, sharey=False, tight_layout=True)

#-------------------------------------------------------------------------RGB
R_DIST = df3['R']
R_mu = np.mean(R_DIST)
R_sig = np.std(R_DIST)
# We can set the number of bins with the *bins* keyword argument.
axes[0].hist(R_DIST, bins=256)
axes[0].set_title('R_dist\n'
fr'$\mu={R_mu:.0f}$, $\sigma={R_sig:.0f}$')

G_DIST = df3['G']
G_mu = np.mean(G_DIST)
G_sig = np.std(G_DIST)
axes[1].hist(G_DIST, bins=256)
axes[1].set_title('G_dist\n'
fr'$\mu={G_mu:.0f}$, $\sigma={G_sig:.0f}$')

B_DIST = df3['B']
B_mu = np.mean(B_DIST)
B_sig = np.std(B_DIST)

fig1, axes = plt.subplots(1, 3, sharey=False, tight_layout=True)

# We can set the number of bins with the *bins* keyword argument.
axes[0].hist(R_DIST, bins=256)
axes[0].set_title('R_dist\n'
fr'$\mu={R_mu:.0f}$, $\sigma={R_sig:.0f}$')

axes[1].hist(G_DIST, bins=256)
axes[1].set_title('G_dist\n'
fr'$\mu={G_mu:.0f}$, $\sigma={G_sig:.0f}$')

axes[2].hist(B_DIST, bins=256)
axes[2].set_title('B_dist\n'
fr'$\mu={B_mu:.0f}$, $\sigma={B_sig:.0f}$')
axes[2].hist(B_DIST, bins=256)
axes[2].set_title('B_dist\n'
fr'$\mu={B_mu:.0f}$, $\sigma={B_sig:.0f}$')

#----------------------------------------------------------------------H Data
#----------------------------------circular mean and stdev since H is a circle
H_DIST = df2['H'].round(0).astype(int)
#H_DIST = H_DIST*2
rads = np.deg2rad(H_DIST)
circmn = circmean(rads*2)
h_mu = np.rad2deg(circmn)
#----------------------------------------------------------------------H Data
#----------------------------------circular mean and stdev since H is a circle
H_DIST = df2['H'].round(0).astype(int)
#H_DIST = H_DIST*2
rads = np.deg2rad(H_DIST)
circmn = circmean(rads*2)
h_mu = np.rad2deg(circmn)


crcstd = circstd(rads*2)
h_sig = np.rad2deg(crcstd)
crcstd = circstd(rads*2)
h_sig = np.rad2deg(crcstd)

#-----------------------------------------------------------------------S Data
S_DIST = df2['S']
s_mu = np.mean(S_DIST)
s_sig = np.std(S_DIST)
#-----------------------------------------------------------------------S Data
S_DIST = df2['S']
s_mu = np.mean(S_DIST)
s_sig = np.std(S_DIST)



#-----------------------------------------------------------------------V Data
V_DIST = df2['V']
#-----------------------------------------------------------------------V Data
V_DIST = df2['V']

#------------For filtering just the V values from the thresholding filter
#FILTERED_V_DIST = [i for i,j in zip(V_DIST, flat_mask) if j == 255]
#------------For filtering just the V values from the thresholding filter
#FILTERED_V_DIST = [i for i,j in zip(V_DIST, flat_mask) if j == 255]

v_mu = np.mean(V_DIST)
v_sig = np.std(V_DIST)
v_mu = np.mean(V_DIST)
v_sig = np.std(V_DIST)
plt.savefig(filename+'_1.png')


fig2, axs = plt.subplots(1, 3, sharey=False, tight_layout=True)
# We can set the number of bins with the *bins* keyword argument.
axs[0].hist(H_DIST, bins=360)
axs[0].set_title('H_dist\n'
fr'$\mu={h_mu:.0f}$, $\sigma={h_sig:.0f}$')

fig2, axs = plt.subplots(1, 3, sharey=False, tight_layout=True)
axs[1].hist(S_DIST, bins=256)
axs[1].set_title('S_dist\n'
fr'$\mu={s_mu:.0f}$, $\sigma={s_sig:.0f}$')

# We can set the number of bins with the *bins* keyword argument.
axs[0].hist(H_DIST, bins=360)
axs[0].set_title('H_dist\n'
fr'$\mu={h_mu:.0f}$, $\sigma={h_sig:.0f}$')
axs[2].hist(V_DIST, bins=256)
axs[2].set_title('V_dist\n'
fr'$\mu={v_mu:.0f}$, $\sigma={v_sig:.0f}$')

axs[1].hist(S_DIST, bins=256)
axs[1].set_title('S_dist\n'
fr'$\mu={s_mu:.0f}$, $\sigma={s_sig:.0f}$')
#-----------------------------------------------------------------------------
#-----------------------------------------------SCATTER PLOT -----------------
#-----------------------------------------------------------------------------

axs[2].hist(V_DIST, bins=256)
axs[2].set_title('V_dist\n'
fr'$\mu={v_mu:.0f}$, $\sigma={v_sig:.0f}$')

#-----------------------------------------------------------------------------
#-----------------------------------------------SCATTER PLOT -----------------
#-----------------------------------------------------------------------------
#----------------------------------------------------------Define size of dots
area = 1.5

#------------------------------------Making sure color values are not squished
graphbounds = pd.Series([0, pi/2, pi])
satbounds = pd.Series([255, 255, 255])

#----------------------------------------------------------Define size of dots
area = 1.5
rads = pd.concat([rads, graphbounds])
S_DIST = pd.concat([S_DIST, satbounds])

#------------------------------------Making sure color values are not squished
graphbounds = pd.Series([0, pi/2, pi])
satbounds = pd.Series([255, 255, 255])
# plt.show()
plt.savefig(filename+'_2.png')
#--------------------------------------------------------------PLOTTING STARTS
fig3 = plt.figure()
axe = fig3.add_subplot(projection='polar')
axe.set_yticks([100, 200, 255])
axe.errorbar(circmn,s_mu,
xerr= crcstd,yerr= s_sig,
capsize=7,
fmt= '^',
c='k')
c = axe.scatter(rads*2,
S_DIST,
c=rads,
s=area,
cmap='hsv',
alpha=1)

rads = pd.concat([rads, graphbounds])
S_DIST = pd.concat([S_DIST, satbounds])

#----------------------------------OpenCV only has H values between 0 and 180
#ax.set_thetamin(0)
#ax.set_thetamax(180)
plt.savefig(filename+'_3.png')

#--------------------------------------------------------------PLOTTING STARTS
fig3 = plt.figure()
axe = fig3.add_subplot(projection='polar')
axe.set_yticks([100, 200, 255])
axe.errorbar(circmn,s_mu,
xerr= crcstd,yerr= s_sig,
capsize=7,
fmt= '^',
c='k')
c = axe.scatter(rads*2,
S_DIST,
c=rads,
s=area,
cmap='hsv',
alpha=1)
return H_DIST, h_mu, h_sig, V_DIST, v_mu, v_sig, S_DIST, s_mu, s_sig


#----------------------------------OpenCV only has H values between 0 and 180
#ax.set_thetamin(0)
#ax.set_thetamax(180)



Expand Down
Loading