-
Notifications
You must be signed in to change notification settings - Fork 1
/
Copy pathevaluate_results.py
196 lines (138 loc) · 5.98 KB
/
evaluate_results.py
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
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
import os, sys, copy, glob, json, time, random, argparse
from shutil import copyfile
from tqdm import tqdm, trange
import mmcv
import imageio
import numpy as np
import torch
import torch.nn as nn
import torch.nn.functional as F
from lib import utils, dvgo
from lib.load_data import load_data
from lib.utils import intersect_sphere
from lib.load_llff import imread
import lib.utils as utils
import cv2
import numpy as np
def load_many_images(global_path,image_dic):
dvg_list = []
for subdir, dirs, files in os.walk(global_path + image_dic):
#print(files)
print(dirs)
list_for_this = []
for file in files:
filepath = subdir + os.sep + file
#print(filepath)
if filepath.endswith(".png"):
image = imread(filepath)/255.
list_for_this.append(image)
# print(image.shape)
# print(filepath)
if list_for_this:
dvg_list.append(list_for_this)
return dvg_list
def get_error_images(loss_func,imgs_dir,target_dir):
score_alls = []
for dir_scene_imgs, dir_scene_target in zip(imgs_dir,target_dir):
score_array = []
#print(len(dir_scene_target))
for img_orig,img_tgt in zip(dir_scene_imgs, dir_scene_target):
#print(img_orig.shape)
score_array.append(loss_func(img_orig,img_tgt))
score_alls.append(round(np.array(score_array).mean(),3))
return np.array(score_alls)
def mask_images(imgs_dir,mask_dir):
masked_img_all = []
for dir_scene_imgs, dir_scene_mask in zip(imgs_dir,mask_dir):
masked_img_scene = []
for img_orig,img_mask in zip(dir_scene_imgs, dir_scene_mask):
masked_img_scene.append(img_orig*img_mask[...,None])
masked_img_all.append(masked_img_scene)
return masked_img_all
def rgb2gray(rgb):
return np.dot(rgb[...,:3], [0.2989, 0.5870, 0.1140])
def create_mask_from_images(global_path,image_dic):
dvg_list = []
for subdir, dirs, files in os.walk(global_path + image_dic):
#print(files)
list_for_this = []
for file in files:
filepath = subdir + os.sep + file
#print(filepath)
if filepath.endswith(".png"):
image = rgb2gray(imread(filepath)/255.)
image[image>0.09] = 1.0
cv2.imwrite(global_path+image_dic+"mask/"+filepath,image)
list_for_this.append(image)
# print(image.shape)
# print(filepath)
if list_for_this:
dvg_list.append(list_for_this)
return dvg_list
print("Helloooo")
#dvg_img = cv2.imread("C:/Users/Temp/Desktop/DirectVoxGO/Results_texts/DirectVoxGo/llff_all_image/africa/000.png")
#gt_img = cv2.imread("C:/Users/Temp/Desktop/DirectVoxGO/Results_texts/Ground_truth/africa/0000.png")
#mine_img = cv2.imread("C:/Users/Temp/Desktop/DirectVoxGO/Results_texts/HashDirectVoxGo++/llff_image_all/africa/000.png")
#plenox_img = cv2.imread("C:/Users/Temp/Desktop/DirectVoxGO/Results_texts/Plenoxels/images_all/africa/0000.png")
#
#
#print(dvg_img.shape)
#print(gt_img.shape)
#print(mine_img.shape)
#print(plenox_img.shape)
#
#img_concat = np.concatenate([dvg_img,plenox_img,mine_img,gt_img],axis=1)
#dvg_img = cv2.imread("C:/Users/Temp/Desktop/DirectVoxGO/Results_texts/HashDirectVoxGo++/llff_foreground/basket/000.png")
#plenox_img = cv2.imread("C:/Users/Temp/Desktop/DirectVoxGO/Results_texts/Plenoxels/images_fg/basket/0000.png")
#
#img_concat = np.concatenate([dvg_img,plenox_img],axis=1)
#dvg_img = cv2.imread("C:/Users/Temp/Desktop/DirectVoxGO/Results_texts/HashDirectVoxGo++/llff_mask/basket/000.png")
#plenox_img = cv2.imread("C:/Users/Temp/Desktop/DirectVoxGO/Results_texts/Plenoxels/mask/basket/0000.png")
#
#img_concat = np.concatenate([dvg_img,plenox_img],axis=1)
#
#
#
#cv2.imwrite("img_test.png",img_concat)
#
#exit(0)
path = "C:/Users/Temp/Desktop/DirectVoxGO/Results_texts/"
dvg_p_mask = create_mask_from_images(path,"HashDirectVoxGo++/llff_foreground/")
plenoxel_dict = create_mask_from_images(path,"Plenoxels/images_fg/")
PSNR = lambda x,y : -10. * np.log10(np.mean(np.square(x - y)))
ssim = lambda x,y :utils.rgb_ssim(x, y, max_val=1)
lpips = lambda x,y :utils.rgb_lpips(np.float32(x), np.float32(y), net_name='vgg', device=torch.device("cpu"))
mask_errors = get_error_images(PSNR,plenoxel_dict,dvg_p_mask)
print("Mask Error:")
print(mask_errors)
mine_dic = load_many_images(path,"HashDirectVoxGo++/llff_image_all/")
gt_dic = load_many_images(path,"Ground_truth/")
plenoxel_dict = load_many_images(path,"Plenoxels/images_all/")
dvg_dic = load_many_images(path,"DirectVoxGo/llff_all_image/")
mask_mine_dic = mask_images(mine_dic,dvg_p_mask)
mask_img_gt = mask_images(gt_dic,dvg_p_mask)
mask_img_plenoxel = mask_images(plenoxel_dict,dvg_p_mask)
mask_img_dvg = mask_images(dvg_dic,dvg_p_mask)
mine_masked_psnr = get_error_images(PSNR,mask_mine_dic,mask_img_gt)
mine_masked_ssim = get_error_images(ssim,mask_mine_dic,mask_img_gt)
mine_masked_lpips = get_error_images(lpips,mask_mine_dic,mask_img_gt)
print("mine:")
print(mine_masked_psnr,mine_masked_psnr.mean())
print(mine_masked_ssim,mine_masked_ssim.mean())
print(mine_masked_lpips,mine_masked_lpips.mean())
print("directvoxgo:")
dvgo_masked_psnr = get_error_images(PSNR,mask_img_dvg,mask_img_gt)
dvgo_masked_ssim = get_error_images(ssim,mask_img_dvg,mask_img_gt)
dvgo_masked_lpips = get_error_images(lpips,mask_img_dvg,mask_img_gt)
print(dvgo_masked_psnr,dvgo_masked_psnr.mean())
print(dvgo_masked_ssim,dvgo_masked_ssim.mean())
print(dvgo_masked_lpips,dvgo_masked_lpips.mean())
plenoxel_masked_psnr = get_error_images(PSNR,mask_img_plenoxel,mask_img_gt)
plenoxel_masked_ssim = get_error_images(ssim,mask_img_plenoxel,mask_img_gt)
plenoxel_masked_lpips = get_error_images(lpips,mask_img_plenoxel,mask_img_gt)
print("plenoxel:")
print(plenoxel_masked_psnr,plenoxel_masked_psnr.mean())
print(plenoxel_masked_ssim,plenoxel_masked_ssim.mean())
print(plenoxel_masked_lpips,plenoxel_masked_lpips.mean())
#print(len(dvg_dic))
#print(len(plenoxel_dict))