File size: 7,941 Bytes
625a17f
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
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
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
import argparse
import json
import tqdm
import cv2
import os
import numpy as np
from pycocotools import mask as mask_utils
import random
from PIL import Image
from natsort import natsorted
from pycocotools.mask import encode, decode, frPyObjects

EVALMODE = "test"


def blend_mask(input_img, binary_mask, alpha=0.5, color="g"):
    if input_img.ndim == 2:
        return input_img
    mask_image = np.zeros(input_img.shape, np.uint8)
    if color == "r":
        mask_image[:, :, 0] = 255
    if color == "g":
        mask_image[:, :, 1] = 255
    if color == "b":
        mask_image[:, :, 2] = 255
    if color == "o":
        mask_image[:, :, 0] = 255
        mask_image[:, :, 1] = 165
        mask_image[:, :, 2] = 0
    if color == "c":
        mask_image[:, :, 0] = 0
        mask_image[:, :, 1] = 255
        mask_image[:, :, 2] = 255
    if color == "p":
        mask_image[:, :, 0] = 128
        mask_image[:, :, 1] = 0
        mask_image[:, :, 2] = 128
    if color == "l":
        mask_image[:, :, 0] = 128
        mask_image[:, :, 1] = 128
        mask_image[:, :, 2] = 0
    if color == "m":
        mask_image[:, :, 0] = 128
        mask_image[:, :, 1] = 128
        mask_image[:, :, 2] = 128
    if color == "q":
        mask_image[:, :, 0] = 165
        mask_image[:, :, 1] = 80
        mask_image[:, :, 2] = 30
    

    mask_image = mask_image * np.repeat(binary_mask[:, :, np.newaxis], 3, axis=2)
    blend_image = input_img[:, :, :].copy()
    pos_idx = binary_mask > 0
    for ind in range(input_img.ndim):
        ch_img1 = input_img[:, :, ind]
        ch_img2 = mask_image[:, :, ind]
        ch_img3 = blend_image[:, :, ind]
        ch_img3[pos_idx] = alpha * ch_img1[pos_idx] + (1 - alpha) * ch_img2[pos_idx]
        blend_image[:, :, ind] = ch_img3
    return blend_image


def upsample_mask(mask, frame):
    H, W = frame.shape[:2]
    mH, mW = mask.shape[:2]

    if W > H:
        ratio = mW / W
        h = H * ratio
        diff = int((mH - h) // 2)
        if diff == 0:
            mask = mask
        else:
            mask = mask[diff:-diff]
    else:
        ratio = mH / H
        w = W * ratio
        diff = int((mW - w) // 2)
        if diff == 0:
            mask = mask
        else:
            mask = mask[:, diff:-diff]

    mask = cv2.resize(mask, (W, H))
    return mask


def downsample(mask, frame):
    H, W = frame.shape[:2]
    mH, mW = mask.shape[:2]

    mask = cv2.resize(mask, (W, H))
    return mask


#datapath /datasegswap
#inference_path /inference_xmem_ego_last/coco
#output /vis_piano
#--show_gt要加上
if __name__ == "__main__":

    color = ['g', 'r', 'b', 'o', 'c', 'p', 'l', 'm', 'q']
    data_path = "/data/work-gcp-europe-west4-a/yuqian_fu/datasets/HANDAL"
    output_path = "/data/work-gcp-europe-west4-a/yuqian_fu/datasets/HANDAL/Handal_vis_results_correct_last"
    mask_base_path = "/data/work-gcp-europe-west4-a/yuqian_fu/datasets/HANDAL/predictions_handal_all"
    json_path = "/data/work-gcp-europe-west4-a/yuqian_fu/datasets/HANDAL/handal_test_all_instruct_correct_videoname.json"


    with open(json_path, "r") as fp:
        datas = json.load(fp)
    print(len(datas))
    
    video_select = ["handal_dataset_fixed_joint_pliers"] # debug
    
    for video_name in tqdm.tqdm(video_select):
        #实验需改动
        data_list = []
        for data in datas:
            if data["video_name"] == video_name:
                data_list.append(data)
        print(len(data_list))
        data_list = random.sample(data_list, 100)
        for data in data_list:
            query_img = cv2.imread(os.path.join(data_path, data['first_frame_image']))
            target_img = cv2.imread(os.path.join(data_path, data['image']))


            # target_gt
            for i,ann in enumerate(data["anns"]):
                mask = decode(ann["segmentation"])
                mask = downsample(mask, target_img)
                out = blend_mask(target_img, mask, color=color[0])
                os.makedirs(
                            f"{output_path}/{video_name}/target_gt", #debug
                            exist_ok=True,
                        )
                img_path1 = data['image']
                tmp_list = img_path1.split("/")[1:]
                joined_path = os.path.join(*tmp_list)
                #print("joined_path:", joined_path)
                output_file_path = os.path.join(output_path, video_name, "target_gt", joined_path)
                #print("output_file_path:", output_file_path)
                dir_path = output_file_path.split("/")[:-1]
                dir_path = "/".join(dir_path)
                os.makedirs(
                            dir_path,  #debug
                            exist_ok=True,)
                cv2.imwrite(
                    output_file_path,  #debug
                    out,
                )
          

            # querty_gt
            for i,ann in enumerate(data["first_frame_anns"]):
                mask = decode(ann["segmentation"])
                mask = downsample(mask, query_img)
                out = blend_mask(query_img, mask, color=color[0])
                # os.makedirs(
                #             f"{output_path}/{video_name}/query_gt",  #debug
                #             exist_ok=True,
                #         )
                new_path_query = data['first_frame_image'].replace(f"{video_name}/", "")
                output_file_path = os.path.join(output_path, video_name, "query_gt", new_path_query)
                #print("new_path_query:", new_path_query)
                #print("query_path:", f"{output_path}/{video_name}/query_gt/{new_path_query}")
                dir_path = output_file_path.split("/")[:-1]
                dir_path = "/".join(dir_path)
                os.makedirs(
                            dir_path,  #debug
                            exist_ok=True,)
                cv2.imwrite(
                    output_file_path,  #debug
                    out,
                )


            # prediction
            mask_path = os.path.join(mask_base_path, data['image'])
            #print(mask_path)
            mask_path = mask_path.replace(".jpg", ".png")
            mask = Image.open(mask_path)
            mask = np.array(mask)
            unique_instances = np.unique(mask)
            unique_instances = unique_instances[unique_instances != 0]
            if len(unique_instances) > 9:
                continue
            for i,instance_value in enumerate(unique_instances):
                binary_mask = (mask == instance_value).astype(np.uint8)
                binary_mask = cv2.resize(binary_mask, (target_img.shape[1], target_img.shape[0]))
                try:
                    binary_mask = upsample_mask(binary_mask, target_img)
                    frame = blend_mask(target_img, binary_mask, color=color[i])
                except:
                    breakpoint()
            
            new_path_predict = data['image'].replace(f"{video_name}/", "")
            #print("new_path_predict:", new_path_predict)
            #print("predict_path:", f"{output_path}/{video_name}/predict/{new_path_predict}")
            output_file_path = os.path.join(output_path, video_name, "predict", new_path_predict)
            dir_path = output_file_path.split("/")[:-1]
            dir_path = "/".join(dir_path)
            os.makedirs(
                        dir_path,  #debug
                        exist_ok=True,)
            cv2.imwrite(
                output_file_path,  #debug
                frame,
            )

            # os.makedirs(
            #                 f"{output_path}/{video_name}/predict",  #debug
            #                 exist_ok=True,
            #             )
            # cv2.imwrite(
            #             f"{output_path}/{video_name}/predict/{new_path_predict}",  #debug
            #             frame,
            #         )