File size: 22,295 Bytes
7acaad7
 
 
 
 
 
 
5069bec
7acaad7
 
 
4a7fc02
7acaad7
 
2507d2f
7acaad7
 
b5957bd
7acaad7
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
7419d98
 
 
 
 
 
 
7acaad7
 
2507d2f
 
 
 
 
 
 
7acaad7
2507d2f
 
 
 
7acaad7
2507d2f
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
7acaad7
2507d2f
 
 
 
 
7acaad7
2507d2f
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
7acaad7
2507d2f
 
 
 
7acaad7
2507d2f
 
7acaad7
2507d2f
 
 
 
 
7acaad7
2507d2f
 
 
 
 
 
 
 
7acaad7
2507d2f
 
7acaad7
2507d2f
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
7acaad7
2507d2f
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
7acaad7
2507d2f
 
 
 
 
 
 
4a7fc02
2507d2f
 
 
 
 
7acaad7
 
2507d2f
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
7acaad7
 
 
 
 
 
2507d2f
 
 
 
 
 
 
 
 
 
7acaad7
 
 
 
3c77caa
7acaad7
2507d2f
 
 
 
 
7acaad7
2507d2f
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
7acaad7
2507d2f
 
 
 
 
 
 
 
 
 
 
7acaad7
2507d2f
 
7acaad7
2507d2f
 
 
 
 
 
 
 
 
 
 
7acaad7
2507d2f
 
 
 
 
 
 
 
 
 
7acaad7
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
2d36d99
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
6ae8c1a
2d36d99
 
 
 
 
 
 
 
10dcc2e
2d36d99
 
 
 
 
 
 
3c77caa
2d36d99
 
 
3c77caa
2d36d99
 
b5957bd
 
 
 
2d36d99
 
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
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
import argparse
import numpy as np
import gradio as gr
from pathlib import Path
from typing import Dict, Any, Optional, Tuple, List, Union
from common.utils import (
    ransac_zoo,
    generate_warp_images,
    load_config,
    get_matcher_zoo,
    run_matching,
    run_ransac,
    gen_examples,
    GRADIO_VERSION,
    ROOT,
)


DESCRIPTION = """
# Image Matching WebUI
This Space demonstrates [Image Matching WebUI](https://github.com/Vincentqyw/image-matching-webui) by vincent qin. Feel free to play with it, or duplicate to run image matching without a queue!
<br/>
πŸ”Ž For more details about supported local features and matchers, please refer to https://github.com/Vincentqyw/image-matching-webui

πŸš€ All algorithms run on CPU for inference, causing slow speeds and high latency. For faster inference, please download the [source code](https://github.com/Vincentqyw/image-matching-webui) for local deployment.

πŸ› Your feedback is valuable to me. Please do not hesitate to report any bugs [here](https://github.com/Vincentqyw/image-matching-webui/issues).
"""


class ImageMatchingApp:
    def __init__(self, server_name="0.0.0.0", server_port=7860, **kwargs):
        self.server_name = server_name
        self.server_port = server_port
        self.config_path = kwargs.get(
            "config", Path(__file__).parent / "config.yaml"
        )
        self.cfg = load_config(self.config_path)
        self.matcher_zoo = get_matcher_zoo(self.cfg["matcher_zoo"])
        self.app = None
        self.init_interface()
        # print all the keys

    def init_matcher_dropdown(self):
        algos = []
        for k, v in self.cfg["matcher_zoo"].items():
            if v.get("enable", True):
                algos.append(k)
        return algos

    def init_interface(self):
        with gr.Blocks() as self.app:
            with gr.Tab("Image Matching"):
                with gr.Row():
                    with gr.Column(scale=1):
                        gr.Image(
                            str(
                                Path(__file__).parent.parent
                                / "assets/logo.webp"
                            ),
                            elem_id="logo-img",
                            show_label=False,
                            show_share_button=False,
                            show_download_button=False,
                        )
                    with gr.Column(scale=3):
                        gr.Markdown(DESCRIPTION)
                with gr.Row(equal_height=False):
                    with gr.Column():
                        with gr.Row():
                            matcher_list = gr.Dropdown(
                                choices=self.init_matcher_dropdown(),
                                value="disk+lightglue",
                                label="Matching Model",
                                interactive=True,
                            )
                            match_image_src = gr.Radio(
                                (
                                    ["upload", "webcam", "clipboard"]
                                    if GRADIO_VERSION > "3"
                                    else ["upload", "webcam", "canvas"]
                                ),
                                label="Image Source",
                                value="upload",
                            )
                        with gr.Row():
                            input_image0 = gr.Image(
                                label="Image 0",
                                type="numpy",
                                image_mode="RGB",
                                height=300 if GRADIO_VERSION > "3" else None,
                                interactive=True,
                            )
                            input_image1 = gr.Image(
                                label="Image 1",
                                type="numpy",
                                image_mode="RGB",
                                height=300 if GRADIO_VERSION > "3" else None,
                                interactive=True,
                            )

                        with gr.Row():
                            button_reset = gr.Button(value="Reset")
                            button_run = gr.Button(
                                value="Run Match", variant="primary"
                            )

                        with gr.Accordion("Advanced Setting", open=False):
                            with gr.Accordion("Matching Setting", open=True):
                                with gr.Row():
                                    match_setting_threshold = gr.Slider(
                                        minimum=0.0,
                                        maximum=1,
                                        step=0.001,
                                        label="Match thres.",
                                        value=0.1,
                                    )
                                    match_setting_max_features = gr.Slider(
                                        minimum=10,
                                        maximum=10000,
                                        step=10,
                                        label="Max features",
                                        value=1000,
                                    )
                                # TODO: add line settings
                                with gr.Row():
                                    detect_keypoints_threshold = gr.Slider(
                                        minimum=0,
                                        maximum=1,
                                        step=0.001,
                                        label="Keypoint thres.",
                                        value=0.015,
                                    )
                                    detect_line_threshold = gr.Slider(
                                        minimum=0.1,
                                        maximum=1,
                                        step=0.01,
                                        label="Line thres.",
                                        value=0.2,
                                    )
                                # matcher_lists = gr.Radio(
                                #     ["NN-mutual", "Dual-Softmax"],
                                #     label="Matcher mode",
                                #     value="NN-mutual",
                                # )
                            with gr.Accordion("RANSAC Setting", open=True):
                                with gr.Row(equal_height=False):
                                    ransac_method = gr.Dropdown(
                                        choices=ransac_zoo.keys(),
                                        value=self.cfg["defaults"][
                                            "ransac_method"
                                        ],
                                        label="RANSAC Method",
                                        interactive=True,
                                    )
                                ransac_reproj_threshold = gr.Slider(
                                    minimum=0.0,
                                    maximum=12,
                                    step=0.01,
                                    label="Ransac Reproj threshold",
                                    value=8.0,
                                )
                                ransac_confidence = gr.Slider(
                                    minimum=0.0,
                                    maximum=1,
                                    step=0.00001,
                                    label="Ransac Confidence",
                                    value=self.cfg["defaults"][
                                        "ransac_confidence"
                                    ],
                                )
                                ransac_max_iter = gr.Slider(
                                    minimum=0.0,
                                    maximum=100000,
                                    step=100,
                                    label="Ransac Iterations",
                                    value=self.cfg["defaults"][
                                        "ransac_max_iter"
                                    ],
                                )
                                button_ransac = gr.Button(
                                    value="Rerun RANSAC", variant="primary"
                                )
                            with gr.Accordion("Geometry Setting", open=False):
                                with gr.Row(equal_height=False):
                                    choice_geometry_type = gr.Radio(
                                        ["Fundamental", "Homography"],
                                        label="Reconstruct Geometry",
                                        value=self.cfg["defaults"][
                                            "setting_geometry"
                                        ],
                                    )

                        # collect inputs
                        state_cache = gr.State({})
                        inputs = [
                            input_image0,
                            input_image1,
                            match_setting_threshold,
                            match_setting_max_features,
                            detect_keypoints_threshold,
                            matcher_list,
                            ransac_method,
                            ransac_reproj_threshold,
                            ransac_confidence,
                            ransac_max_iter,
                            choice_geometry_type,
                            gr.State(self.matcher_zoo),
                            # state_cache,
                        ]

                        # Add some examples
                        with gr.Row():
                            # Example inputs
                            gr.Examples(
                                examples=gen_examples(),
                                inputs=inputs,
                                outputs=[],
                                fn=run_matching,
                                cache_examples=False,
                                label=(
                                    "Examples (click one of the images below to Run"
                                    " Match). Thx: WxBS"
                                ),
                            )
                        with gr.Accordion("Supported Algorithms", open=False):
                            # add a table of supported algorithms
                            self.display_supported_algorithms()

                    with gr.Column():
                        output_keypoints = gr.Image(
                            label="Keypoints", type="numpy"
                        )
                        output_matches_raw = gr.Image(
                            label="Raw Matches",
                            type="numpy",
                        )
                        output_matches_ransac = gr.Image(
                            label="Ransac Matches", type="numpy"
                        )
                        with gr.Accordion(
                            "Open for More: Matches Statistics", open=False
                        ):
                            matches_result_info = gr.JSON(
                                label="Matches Statistics"
                            )
                            matcher_info = gr.JSON(label="Match info")

                        with gr.Accordion(
                            "Open for More: Warped Image", open=False
                        ):
                            output_wrapped = gr.Image(
                                label="Wrapped Pair", type="numpy"
                            )
                            with gr.Accordion(
                                "Open for More: Geometry info", open=False
                            ):
                                geometry_result = gr.JSON(
                                    label="Reconstructed Geometry"
                                )

                    # callbacks
                    match_image_src.change(
                        fn=self.ui_change_imagebox,
                        inputs=match_image_src,
                        outputs=input_image0,
                    )
                    match_image_src.change(
                        fn=self.ui_change_imagebox,
                        inputs=match_image_src,
                        outputs=input_image1,
                    )

                    # collect outputs
                    outputs = [
                        output_keypoints,
                        output_matches_raw,
                        output_matches_ransac,
                        matches_result_info,
                        matcher_info,
                        geometry_result,
                        output_wrapped,
                        state_cache,
                    ]
                    # button callbacks
                    button_run.click(
                        fn=run_matching, inputs=inputs, outputs=outputs
                    )

                    # Reset images
                    reset_outputs = [
                        input_image0,
                        input_image1,
                        match_setting_threshold,
                        match_setting_max_features,
                        detect_keypoints_threshold,
                        matcher_list,
                        input_image0,
                        input_image1,
                        match_image_src,
                        output_keypoints,
                        output_matches_raw,
                        output_matches_ransac,
                        matches_result_info,
                        matcher_info,
                        output_wrapped,
                        geometry_result,
                        ransac_method,
                        ransac_reproj_threshold,
                        ransac_confidence,
                        ransac_max_iter,
                        choice_geometry_type,
                    ]
                    button_reset.click(
                        fn=self.ui_reset_state,
                        inputs=None,
                        outputs=reset_outputs,
                    )

                    # run ransac button action
                    button_ransac.click(
                        fn=run_ransac,
                        inputs=[
                            state_cache,
                            choice_geometry_type,
                            ransac_method,
                            ransac_reproj_threshold,
                            ransac_confidence,
                            ransac_max_iter,
                        ],
                        outputs=[
                            output_matches_ransac,
                            matches_result_info,
                            output_wrapped,
                        ],
                    )

                    # estimate geo
                    choice_geometry_type.change(
                        fn=generate_warp_images,
                        inputs=[
                            input_image0,
                            input_image1,
                            geometry_result,
                            choice_geometry_type,
                        ],
                        outputs=[output_wrapped, geometry_result],
                    )
            with gr.Tab("Under construction"):
                self.init_tab_sfm()

    def init_tab_sfm(self):
        with gr.Row():
            with gr.Column():
                with gr.Row():
                    gr.Textbox("Under construction", label="A", visible=True)
                    gr.Textbox("Under construction", label="B", visible=True)
                    gr.Textbox("Under construction", label="C", visible=True)
                with gr.Row():
                    with gr.Accordion("Open for More", open=False):
                        gr.Textbox(
                            "Under construction", label="A1", visible=True
                        )
                        gr.Textbox(
                            "Under construction", label="B1", visible=True
                        )
                        gr.Textbox(
                            "Under construction", label="C1", visible=True
                        )
            with gr.Column():
                gr.Textbox("Under construction", label="D", visible=True)
                gr.Textbox("Under construction", label="E", visible=True)
                gr.Textbox("Under construction", label="F", visible=True)

    def run(self):
        self.app.queue().launch(
            server_name=self.server_name,
            server_port=self.server_port,
            share=False,
        )

    def ui_change_imagebox(self, choice):
        """
        Updates the image box with the given choice.

        Args:
            choice (list): The list of image sources to be displayed in the image box.

        Returns:
            dict: A dictionary containing the updated value, sources, and type for the image box.
        """
        ret_dict = {
            "value": None,  # The updated value of the image box
            "__type__": "update",  # The type of update for the image box
        }
        if GRADIO_VERSION > "3":
            return {
                **ret_dict,
                "sources": choice,  # The list of image sources to be displayed
            }
        else:
            return {
                **ret_dict,
                "source": choice,  # The list of image sources to be displayed
            }

    def ui_reset_state(
        self,
        *args: Any,
    ) -> Tuple[
        Optional[np.ndarray],
        Optional[np.ndarray],
        float,
        int,
        float,
        str,
        Dict[str, Any],
        Dict[str, Any],
        str,
        Optional[np.ndarray],
        Optional[np.ndarray],
        Optional[np.ndarray],
        Dict[str, Any],
        Dict[str, Any],
        Optional[np.ndarray],
        Dict[str, Any],
        str,
        int,
        float,
        int,
    ]:
        """
        Reset the state of the UI.

        Returns:
            tuple: A tuple containing the initial values for the UI state.
        """
        key: str = list(self.matcher_zoo.keys())[
            0
        ]  # Get the first key from matcher_zoo
        return (
            None,  # image0: Optional[np.ndarray]
            None,  # image1: Optional[np.ndarray]
            self.cfg["defaults"][
                "match_threshold"
            ],  # matching_threshold: float
            self.cfg["defaults"]["max_keypoints"],  # max_features: int
            self.cfg["defaults"][
                "keypoint_threshold"
            ],  # keypoint_threshold: float
            key,  # matcher: str
            self.ui_change_imagebox("upload"),  # input image0: Dict[str, Any]
            self.ui_change_imagebox("upload"),  # input image1: Dict[str, Any]
            "upload",  # match_image_src: str
            None,  # keypoints: Optional[np.ndarray]
            None,  # raw matches: Optional[np.ndarray]
            None,  # ransac matches: Optional[np.ndarray]
            {},  # matches result info: Dict[str, Any]
            {},  # matcher config: Dict[str, Any]
            None,  # warped image: Optional[np.ndarray]
            {},  # geometry result: Dict[str, Any]
            self.cfg["defaults"]["ransac_method"],  # ransac_method: str
            self.cfg["defaults"][
                "ransac_reproj_threshold"
            ],  # ransac_reproj_threshold: float
            self.cfg["defaults"][
                "ransac_confidence"
            ],  # ransac_confidence: float
            self.cfg["defaults"]["ransac_max_iter"],  # ransac_max_iter: int
            self.cfg["defaults"]["setting_geometry"],  # geometry: str
        )

    def display_supported_algorithms(self, style="tab"):
        def get_link(link, tag="Link"):
            return "[{}]({})".format(tag, link) if link is not None else "None"

        data = []
        cfg = self.cfg["matcher_zoo"]
        if style == "md":
            markdown_table = "| Algo. | Conference | Code | Project | Paper |\n"
            markdown_table += (
                "| ----- | ---------- | ---- | ------- | ----- |\n"
            )

            for k, v in cfg.items():
                if not v["info"]["display"]:
                    continue
                github_link = get_link(v["info"]["github"])
                project_link = get_link(v["info"]["project"])
                paper_link = get_link(
                    v["info"]["paper"],
                    (
                        Path(v["info"]["paper"]).name[-10:]
                        if v["info"]["paper"] is not None
                        else "Link"
                    ),
                )

                markdown_table += "{}|{}|{}|{}|{}\n".format(
                    v["info"]["name"],  # display name
                    v["info"]["source"],
                    github_link,
                    project_link,
                    paper_link,
                )
            return gr.Markdown(markdown_table)
        elif style == "tab":
            for k, v in cfg.items():
                if not v["info"].get("display", True):
                    continue
                data.append(
                    [
                        v["info"]["name"],
                        v["info"]["source"],
                        v["info"]["github"],
                        v["info"]["paper"],
                        v["info"]["project"],
                    ]
                )
            tab = gr.Dataframe(
                headers=["Algo.", "Conference", "Code", "Paper", "Project"],
                datatype=["str", "str", "str", "str", "str"],
                col_count=(5, "fixed"),
                value=data,
                # wrap=True,
                # min_width = 1000,
                # height=1000,
            )
            return tab