从坐标到选择单个斑点/对象

时间:2018-12-20 17:38:04

标签: python image cross-correlation

我已经成功地识别出图像中的特定对象,现在我想进行标记,然后将这些对象转移到新图像中以进行进一步处理。但是,我对如何实现这一点有些困惑。这是我使用的脚本。

coordinates = peak_local_max(rack_matches, min_distance=100)

coordinates = coordinates[np.argsort(rack_matches[coordinates[:,0],coordinates[:,1]])[-8:],:]

L = skime.label(imbw_open)

R = skime.regionprops(L)
rack_regions = []
for r in range(len(R)):
    for c in range(8):
        if np.all(coordinates[c,:]) in R[r]['Coordinates']:
            print(coordinates[c,:])
            rack_regions.append(region)

racks_img = np.zeros(shape(L))
for region in rack_regions:
    racks_img[region['Coordinates'][:,0],region['Coordinates'][:,1]] = 1

coords_img = np.zeros(shape(L))
coords_img[coordinates[:,0], coordinates[:,1]] = 1
coords_img = skimo.binary_dilation(coords_img, skimo.disk(3))

0 个答案:

没有答案