OpenSourceIrisRecognition/methods/HDBIF/Python/hdbif.py at main · CVRL/OpenSourceIrisRecognition · GitHub

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
from modules.irisRecognition import irisRecognition
from modules.utils import get_cfg
import argparse
import glob
from PIL import Image
import os
import cv2
import numpy as np

def main(cfg):

    irisRec = irisRecognition(cfg)

    if not os.path.exists('./dataProcessed/'):
        os.mkdir('./dataProcessed/')

    if not os.path.exists('./templates/'):
        os.mkdir('./templates/')

    # Get the list of images to process
    filename_list = []
    image_list = []
    extensions = ["bmp", "png", "gif", "jpg", "jpeg", "tiff", "tif"]
    for ext in extensions:
        for filename in glob.glob("./data/*." + ext):
            im = Image.fromarray(np.array(Image.open(filename).convert("RGB"))[:, :, 0], "L")
            image_list.append(im)
            filename_list.append(os.path.basename(filename))

    # Segmentation, normalization and encoding
    polar_mask_list = []
    code_list = []
    for im,fn in zip(image_list,filename_list):

        print(fn)

        # convert to ISO-compliant aspect ratio (4:3) and resize to ISO-compliant resolution: 640x480
        im = irisRec.fix_image(im)

        # segmentation mask and circular approximation:
        mask, pupil_xyr, iris_xyr = irisRec.segment_and_circApprox(im)
        im_mask = Image.fromarray(np.where(mask > 0.5, 255, 0).astype(np.uint8), 'L')

        # cartesian to polar transformation:
        im_polar, mask_polar = irisRec.cartToPol_torch(im, mask, pupil_xyr, iris_xyr)
        polar_mask_list.append(mask_polar)

        # human-driven BSIF encoding:
        code = irisRec.extractCode(im_polar)
        #print(code.shape)
        code_list.append(code)

        # DEBUG: save selected processing results
        im_mask.save("./dataProcessed/" + os.path.splitext(fn)[0] + "_seg_mask.png")
        imVis = irisRec.segmentVis(im,mask,pupil_xyr,iris_xyr)
        path = "./dataProcessed/" + os.path.splitext(fn)[0]
        cv2.imwrite(path + "_seg_vis.png",imVis)
        cv2.imwrite(path + "_im_polar.png",im_polar)
        cv2.imwrite(path + "_mask_polar.png",mask_polar)
        np.savez_compressed("./templates/" + os.path.splitext(fn)[0] + "_tmpl.npz",code)
        for singleFilterCode, filter_size in zip(code, irisRec.filter_sizes):
            for i in range(singleFilterCode.shape[0]):
                cv2.imwrite(("%s_code_filter%dx%d_%d.png" % (path,filter_size,filter_size,i)),255*singleFilterCode[i,:,:])

    # Matching (all-vs-all, as an example)
    for code1,mask1,fn1,i in zip(code_list,polar_mask_list,filename_list,range(len(code_list))):
        for code2,mask2,fn2,j in zip(code_list,polar_mask_list,filename_list,range(len(code_list))):
            if i < j:
                score, shift = irisRec.matchCodesEfficient(code1, code2, mask1, mask2)
                print("{} <-> {} : {:.3f} (mutual rot: {:.2f} deg)".format(fn1,fn2,score,360*shift/irisRec.polar_width))

    return None

if __name__ == "__main__":
    parser = argparse.ArgumentParser()
    parser.add_argument("--cfg_path",
                        type=str,
                        default="cfg_baseline.yaml",
                        help="path of the configuration file")
    args = parser.parse_args()
    main(get_cfg(args.cfg_path))