104     img = numpy.ndarray(shape=(2*h*w*4,), dtype='<f', buffer=cap["data"])
105     analysis_image = img.reshape(2,h,w,4)
132 def unpack_raw10_image(img):  argument
144     if img.shape[1] % 5 != 0:
146     w = img.shape[1]*4/5
147     h = img.shape[0]
149     msbs = numpy.delete(img, numpy.s_[4::5], 1)
154     lsbs = img[::, 4::5].reshape(h,w/4)
186 def unpack_raw12_image(img):  argument
198     if img.shape[1] % 3 != 0:
200     w = img.shape[1]*2/3
201     h = img.shape[0]
203     msbs = numpy.delete(img, numpy.s_[2::3], 1)
208     lsbs = img[::, 2::3].reshape(h,w/2)
269         img = numpy.ndarray(shape=(h*w,), dtype='<u2',
271         img = img.astype(numpy.float32).reshape(h,w) / white_level
293                 img = img[ycrop:ycrop+hcrop,xcrop:xcrop+wcrop]
302         imgs = [img[::2].reshape(w*h/2)[::2].reshape(h/2,w/2,1),
303                 img[::2].reshape(w*h/2)[1::2].reshape(h/2,w/2,1),
304                 img[1::2].reshape(w*h/2)[::2].reshape(h/2,w/2,1),
305                 img[1::2].reshape(w*h/2)[1::2].reshape(h/2,w/2,1)]
414     img = numpy.dstack([r_plane,(gr_plane+gb_plane)/2.0,b_plane])
415     img = (((img.reshape(h,w,3) - black_levels) * scale) * gains).clip(0.0,1.0)
416     img = numpy.dot(img.reshape(w*h,3), ccm.T).reshape(h,w,3).clip(0.0,1.0)
417     return img
500     img = Image.open(fname)
501     w = img.size[0]
502     h = img.size[1]
503     a = numpy.array(img)
586     img = Image.open(cStringIO.StringIO(jpeg_buffer))
587     w = img.size[0]
588     h = img.size[1]
589     return numpy.array(img).reshape(h,w,3) / 255.0
592 def apply_lut_to_image(img, lut):  argument
623     return (lut[(img * m).astype(numpy.uint16)] / m).astype(numpy.float32)
626 def apply_matrix_to_image(img, mat):  argument
643     h = img.shape[0]
644     w = img.shape[1]
646     img2.reshape(w*h*3)[:] = (numpy.dot(img.reshape(h*w, 3), mat.T)
651 def get_image_patch(img, xnorm, ynorm, wnorm, hnorm):  argument
661     hfull = img.shape[0]
662     wfull = img.shape[1]
667     if len(img.shape)==2:
668         return img[ytile:ytile+htile,xtile:xtile+wtile].copy()
670         return img[ytile:ytile+htile,xtile:xtile+wtile,:].copy()
673 def compute_image_means(img):  argument
683     chans = img.shape[2]
685         means.append(numpy.mean(img[:,:,i], dtype=numpy.float64))
689 def compute_image_variances(img):  argument
699     chans = img.shape[2]
701         variances.append(numpy.var(img[:,:,i], dtype=numpy.float64))
705 def compute_image_snrs(img):  argument
714     means = compute_image_means(img)
715     variances = compute_image_variances(img)
721 def compute_image_max_gradients(img):  argument
731     chans = img.shape[2]
733         grads.append(numpy.amax(numpy.gradient(img[:, :, i])))
737 def write_image(img, fname, apply_gamma=False):  argument
755         img = apply_lut_to_image(img, DEFAULT_GAMMA_LUT)
756     (h, w, chans) = img.shape
758         Image.fromarray((img * 255.0).astype(numpy.uint8), "RGB").save(fname)
760         img3 = (img * 255.0).astype(numpy.uint8).repeat(3).reshape(h,w,3)
766 def downscale_image(img, f):  argument
784     h,w,chans = img.shape
789     img = img[0:h:,0:w:,::]
792         ch = img.reshape(h*w*chans)[i::chans].reshape(h,w)
796     img = numpy.vstack(chs).T.reshape(h/f,w/f,chans)
797     return img
800 def compute_image_sharpness(img):  argument
810     chans = img.shape[2]
813         luma = img[:, :, 0]
815         luma = 0.299 * img[:,:,0] + 0.587 * img[:,:,1] + 0.114 * img[:,:,2]
821 def normalize_img(img):  argument
829     return (img - numpy.amin(img))/(numpy.amax(img) - numpy.amin(img))
850 def rotate_img_per_argv(img):  argument
858     img_out = img