valentinh
/
geetools_VH
forked from kilianv/CoastSat_WRL
1
0
Fork 0
Code Issues Pull Requests Releases Wiki Activity

updated sand classification

Browse Source 
if buffer_size = False, then classify the entire image
development
kvos 7 years ago
parent c31ae42fbd
commit 35f421bfcb
2 changed files with 56 additions and 45 deletions
Show Stats Download Patch File Download Diff File

7
functions/sds.py
Unescape Escape View File

@ -610,6 +610,7 @@ def convert_epsg(points, epsg_in, epsg_out):
def classify_sand_unsupervised(im_ms_ps, im_pan, cloud_mask, wl_pix, buffer_size, min_beach_size, plot_bool): def classify_sand_unsupervised(im_ms_ps, im_pan, cloud_mask, wl_pix, buffer_size, min_beach_size, plot_bool):
""" """
Classifies sand pixels using an unsupervised algorithm (Kmeans) Classifies sand pixels using an unsupervised algorithm (Kmeans)
Set buffer size to False if you
KV WRL 2018 KV WRL 2018
@ -623,8 +624,9 @@ def classify_sand_unsupervised(im_ms_ps, im_pan, cloud_mask, wl_pix, buffer_size
2D cloud mask with True where cloud pixels are 2D cloud mask with True where cloud pixels are
wl_pix: list of np.ndarray wl_pix: list of np.ndarray
list of arrays containig the pixel coordinates of the water line list of arrays containig the pixel coordinates of the water line
buffer_size: int buffer_size: int or False
radius of the disk used to create a buffer around the water line radius of the disk used to create a buffer around the water line
when False, the entire image is considered for kmeans
min_beach_size: int min_beach_size: int
minimum number of connected pixels belonging to a single beach minimum number of connected pixels belonging to a single beach
plot_bool: boolean plot_bool: boolean
@ -644,6 +646,7 @@ def classify_sand_unsupervised(im_ms_ps, im_pan, cloud_mask, wl_pix, buffer_size
vec_features[:,[0,1,2,3]] = vec_ms_ps[:,[0,1,2,3]] vec_features[:,[0,1,2,3]] = vec_ms_ps[:,[0,1,2,3]]
vec_features[:,4] = vec_pan vec_features[:,4] = vec_pan
if buffer_size:
# create binary image with ones where the detected water lines is # create binary image with ones where the detected water lines is
im_buffer = np.zeros((im_ms_ps.shape[0], im_ms_ps.shape[1])) im_buffer = np.zeros((im_ms_ps.shape[0], im_ms_ps.shape[1]))
for i, contour in enumerate(wl_pix): for i, contour in enumerate(wl_pix):
@ -654,6 +657,8 @@ def classify_sand_unsupervised(im_ms_ps, im_pan, cloud_mask, wl_pix, buffer_size
se = morphology.disk(buffer_size) se = morphology.disk(buffer_size)
im_buffer = morphology.binary_dilation(im_buffer, se) im_buffer = morphology.binary_dilation(im_buffer, se)
vec_buffer = (im_buffer == 1).reshape(im_ms_ps.shape[0] * im_ms_ps.shape[1]) vec_buffer = (im_buffer == 1).reshape(im_ms_ps.shape[0] * im_ms_ps.shape[1])
else:
vec_buffer = np.ones((vec_pan.shape[0]))
# add cloud mask to buffer # add cloud mask to buffer
vec_buffer= np.logical_and(vec_buffer, ~vec_mask) vec_buffer= np.logical_and(vec_buffer, ~vec_mask)
# perform kmeans (6 clusters) # perform kmeans (6 clusters)

74
read_images.py
Unescape Escape View File

@ -40,6 +40,8 @@ plot_bool = False # if you want the plots
prob_high = 99.9 # upper probability to clip and rescale pixel intensity prob_high = 99.9 # upper probability to clip and rescale pixel intensity
min_contour_points = 100# minimum number of points contained in each water line min_contour_points = 100# minimum number of points contained in each water line
output_epsg = 28356 # GDA94 / MGA Zone 56 output_epsg = 28356 # GDA94 / MGA Zone 56
buffer_size = 10 # radius (in pixels) of disk for buffer (pixel classification)
min_beach_size = 50 # number of pixels in a beach (pixel classification)
# load metadata (timestamps and epsg code) for the collection # load metadata (timestamps and epsg code) for the collection
satname = 'L8' satname = 'L8'
@ -127,7 +129,7 @@ for i in range(N):
idx_skipped.append(i) idx_skipped.append(i)
continue continue
else: else:
del shorelines[idx_samedate] # del shorelines[idx_samedate]
del t[idx_samedate] del t[idx_samedate]
del cloud_cover_ts[idx_samedate] del cloud_cover_ts[idx_samedate]
del date_acquired_ts[idx_samedate] del date_acquired_ts[idx_samedate]
@ -150,39 +152,43 @@ for i in range(N):
# convert to output epsg spatial reference # convert to output epsg spatial reference
wl = sds.convert_epsg(wl_coords, input_epsg, output_epsg) wl = sds.convert_epsg(wl_coords, input_epsg, output_epsg)
# plot a figure to select the correct water line and discard cloudy images # classify sand pixels
plt.figure() im_sand = sds.classify_sand_unsupervised(im_ms_ps, im_pan, cloud_mask, wl_pix, False, min_beach_size, True)
cmap = cm.get_cmap('jet')
plt.subplot(121) # # plot a figure to select the correct water line and discard cloudy images
plt.imshow(im_ms_ps[:,:,[2,1,0]]) # plt.figure()
for j,contour in enumerate(wl_pix): # cmap = cm.get_cmap('jet')
colours = cmap(np.linspace(0, 1, num=len(wl_pix))) # plt.subplot(121)
plt.plot(contour[:, 1], contour[:, 0], linewidth=2, color=colours[j,:]) # plt.imshow(im_ms_ps[:,:,[2,1,0]])
plt.axis('image') # for j,contour in enumerate(wl_pix):
plt.title(file_names_pan[i]) # colours = cmap(np.linspace(0, 1, num=len(wl_pix)))
plt.subplot(122) # plt.plot(contour[:, 1], contour[:, 0], linewidth=2, color=colours[j,:])
centroids = [] # plt.axis('image')
for j,contour in enumerate(wl): # plt.title(file_names_pan[i])
colours = cmap(np.linspace(0, 1, num=len(wl))) # plt.subplot(122)
centroids.append([np.mean(contour[:, 0]),np.mean(contour[:, 1])]) # centroids = []
plt.plot(contour[:, 0], contour[:, 1], linewidth=2, color=colours[j,:]) # for j,contour in enumerate(wl):
plt.plot(np.mean(contour[:, 0]), np.mean(contour[:, 1]), 'o', color=colours[j,:]) # colours = cmap(np.linspace(0, 1, num=len(wl)))
plt.plot(refpoints[:,0], refpoints[:,1], 'k.') # centroids.append([np.mean(contour[:, 0]),np.mean(contour[:, 1])])
plt.axis('equal') # plt.plot(contour[:, 0], contour[:, 1], linewidth=2, color=colours[j,:])
plt.title(file_names_pan[i]) # plt.plot(np.mean(contour[:, 0]), np.mean(contour[:, 1]), 'o', color=colours[j,:])
mng = plt.get_current_fig_manager() # plt.plot(refpoints[:,0], refpoints[:,1], 'k.')
mng.window.showMaximized() # plt.axis('equal')
plt.tight_layout() # plt.title(file_names_pan[i])
plt.draw() # mng = plt.get_current_fig_manager()
# click on the left image to discard, otherwise on the closest centroid in the right image # mng.window.showMaximized()
pt_in = np.array(ginput(n=1, timeout=1000)) # plt.tight_layout()
if pt_in[0][0] < 10000: # plt.draw()
print('skip ' + str(i) + ' - manual') # # click on the left image to discard, otherwise on the closest centroid in the right image
idx_skipped.append(i) # pt_in = np.array(ginput(n=1, timeout=1000))
continue # if pt_in[0][0] < 10000:
# get contour that was selected (clock closest to centroid) # print('skip ' + str(i) + ' - manual')
dist_centroid = [np.linalg.norm(_ - pt_in) for _ in centroids] # idx_skipped.append(i)
shorelines.append(wl[np.argmin(dist_centroid)]) # continue
# # get contour that was selected (click closest to centroid)
# dist_centroid = [np.linalg.norm(_ - pt_in) for _ in centroids]
# shorelines.append(wl[np.argmin(dist_centroid)])
t.append(timestamps_sorted[i]) t.append(timestamps_sorted[i])
cloud_cover_ts.append(cloud_cover) cloud_cover_ts.append(cloud_cover)
acc_georef_ts.append(acc_georef_sorted[i]) acc_georef_ts.append(acc_georef_sorted[i])

Write Preview
Loading…
Cancel
Save