def serve_mosaic_image(time_series=None, pid=None, params='/'):
"""Generate a mosaic of ROIs from a sample bin.
params include the following, with default values
- series (mvco) - time series (FIXME: handle with resolver, clarify difference between namespace, time series, pid, lid)
- size (1024x1024) - size of the mosaic image
- page (1) - page. for typical image sizes the entire bin does not fit and so is split into pages.
- scale - scaling factor for image dimensions """
# parse params
params = parse_params(params, size='1024x1024',page=1,scale=1.0)
(w,h) = tuple(map(int,re.split('x',params[SIZE])))
scale = float(params[SCALE])
page = int(params[PAGE])
# parse pid/lid
hit = pid_resolver.resolve(pid=pid)
# perform layout operation
scaled_size = (int(w/scale), int(h/scale))
layout = get_mosaic_layout(hit.bin_pid, scaled_size, page)
# serve JSON on request
if hit.extension == 'json':
return jsonr(list(layout2json(layout, scale)))
# resolve ROI file
roi_path = resolve_roi(hit.bin_pid)
# read all images needed for compositing and inject into Tiles
with open(roi_path,'rb') as roi_file:
for tile in layout:
target = tile.image
# FIXME use fast stitching
tile.image = as_pil(get_fast_stitched_roi(hit.bin_pid, target[TARGET_NUMBER]))
# produce and serve composite image
mosaic_image = thumbnail(mosaic.composite(layout, scaled_size, mode='L', bgcolor=160), (w,h))
(pil_format, mimetype) = image_types(hit)
return image_response(mosaic_image, pil_format, mimetype)
def im2bytes(im):
"""Convert a numpy uint8 array to a bytearray"""
im = as_pil(im)
buf = BytesIO()
with tempfile.SpooledTemporaryFile() as imtemp:
im.save(imtemp,'PNG')
imtemp.seek(0)
shutil.copyfileobj(imtemp, buf)
return buf.getvalue()
def stitch_raw(targets,images,box=None,background=0):
# compute bounds relative to the camera field
if box is None:
box = stitched_box(targets)
(x,y,w,h) = box
# now we swap width and height to rotate the image 90 degrees
s = Image.new('L',(h,w),background) # the stitched image with a missing region
for (roi,image) in zip(targets,images):
image = as_pil(image)
rx = roi[LEFT] - x
ry = roi[BOTTOM] - y
rw = roi[WIDTH]
rh = roi[HEIGHT]
roi_box = (ry, rx, ry + rh, rx + rw)
s.paste(image, roi_box) # paste in the right location
return np.array(s)
def image_response(image,format,mimetype):
"""Construct a Flask Response object for the given image, PIL format, and MIME type."""
buf = StringIO()
im = as_pil(image).save(buf,format)
return Response(buf.getvalue(), mimetype=mimetype)
def bright_mv(image,mask=None):
mask = as_pil(mask)
eh = image.histogram(mask)
# toast extrema
return bright_mv_hist(eh)
def stitch(targets,images):
# compute bounds relative to the camera field
(x,y,w,h) = stitched_box(targets)
# note that w and h are switched from here on out to rotate 90 degrees.
# step 1: compute masks
s = as_pil(stitch_raw(targets,images,(x,y,w,h))) # stitched ROI's with black gaps
rois_mask = as_pil(mask(targets)) # a mask of where the ROI's are
gaps_mask = ImageChops.invert(rois_mask) # its inverse is where the gaps are
edges = edges_mask(targets,images) # edges are pixels along the ROI edges
# step 2: estimate background from edges
# compute the mean and variance of the edges
(mean,variance) = bright_mv(s,edges)
# now use that as an estimated background
flat_bg = Image.new('L',(h,w),mean) # FIXME
s.paste(mean,None,gaps_mask)
# step 3: compute "probable background": low luminance delta from estimated bg
bg = extract_background(s,flat_bg)
# also mask out the gaps, which are not "probable background"
bg.paste(255,None,gaps_mask)
# step 3a: improve mean/variance estimate
(mean,variance) = bright_mv(bg)
std_dev = sqrt(variance)
# step 4: sample probable background to compute RBF for illumination gradient
# grid
div = 6
means = []
nodes = []
rad = avg([h,w]) / div
rad_step = int(rad/2)+1
for x in range(0,h+rad,rad):
for y in range(0,w+rad,rad):
for r in range(rad,max(h,w),int(rad/3)+1):
box = (max(0,x-r),max(0,y-r),min(h-1,x+r),min(w-1,y+r))
region = bg.crop(box)
nabe = region.histogram()
(m,v) = bright_mv_hist(nabe)
if m > 0 and m < 255: # reject outliers
nodes.append((x,y))
means.append(m)
break
# now construct radial basis functions for mean, based on the samples
mean_rbf = interpolate.Rbf([x for x,y in nodes], [y for x,y in nodes], means, epsilon=rad)
# step 5: fill gaps with mean based on RBF and variance from bright_mv(edges)
mask_pix = gaps_mask.load()
noise = Image.new('L',(h,w),mean)
noise_pix = noise.load()
np.random.seed(0)
gaussian = np.random.normal(0, 1.0, size=(h,w)) # it's normal
std_dev *= 0.66 # err on the side of smoother rather than noisier
mask_x = []
mask_y = []
for x in xrange(h):
for y in xrange(w):
if mask_pix[x,y] == 255: # only for pixels in the mask
mask_x.append(x)
mask_y.append(y)
rbf_fill = mean_rbf(np.array(mask_x), np.array(mask_y))
for x,y,r in zip(mask_x, mask_y, rbf_fill):
# fill is illumination gradient + noise
noise_pix[x,y] = r + (gaussian[x,y] * std_dev)
# step 6: final composite
s.paste(noise,None,gaps_mask)
return (np.array(s),rois_mask)
