我需要通过非整数因子(例如,100x100阵列到45x45阵列)对2D numpy数组进行下采样,执行局部平均,就像Photoshop / gimp会对图像执行此操作一样。我需要双精度。目前的选择不能很好。
scipy.ndimage.zoom
不执行平均,并且基本上使用
最近邻采样(见上一个问题scipy.ndimage.interpolation.zoom uses nearest-neighbor-like algorithm for scaling-down)
scipy.misc.imresize
将数组转换为int8;我需要更多
精度和浮点
skimage.transform.rescale
也使用最近邻居并将您转发到skimage.transform.downscale_local_mean
进行本地平均,
skimage.transform.downscale_local_mean
只能执行整数缩放因子(如果因子是非整数,则使用零填充图像)。整数缩放因子是一个微不足道的numpy excersice。
我是否错过了其他选择?
答案 0 :(得分:1)
我最后编写了一个小函数,使用scipy.ndimage.zoom
升级图像,但是为了缩小尺寸,它首先将其升级为原始形状的倍数,然后通过块平均缩小。它接受scipy.zoom
(order
和prefilter
)
我仍然在寻找使用可用软件包的清洁解决方案。
def zoomArray(inArray, finalShape, sameSum=False, **zoomKwargs):
inArray = np.asarray(inArray, dtype = np.double)
inShape = inArray.shape
assert len(inShape) == len(finalShape)
mults = []
for i in range(len(inShape)):
if finalShape[i] < inShape[i]:
mults.append(int(np.ceil(inShape[i]/finalShape[i])))
else:
mults.append(1)
tempShape = tuple([i * j for i,j in zip(finalShape, mults)])
zoomMultipliers = np.array(tempShape) / np.array(inShape) + 0.0000001
rescaled = zoom(inArray, zoomMultipliers, **zoomKwargs)
for ind, mult in enumerate(mults):
if mult != 1:
sh = list(rescaled.shape)
assert sh[ind] % mult == 0
newshape = sh[:ind] + [sh[ind] / mult, mult] + sh[ind+1:]
rescaled.shape = newshape
rescaled = np.mean(rescaled, axis = ind+1)
assert rescaled.shape == finalShape
if sameSum:
extraSize = np.prod(finalShape) / np.prod(inShape)
rescaled /= extraSize
return rescaled