File size: 7,637 Bytes
796e051
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
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
import os, sys, math
from typing import List, Optional, Tuple, Dict
import numpy as np
from PIL import Image, ImageDraw, ImageFont




PALETTE: List[Tuple[int, int, int]] = [
    (255, 80, 80),
    (80, 255, 80),
    (80, 160, 255),
    (255, 200, 80),
    (200, 80, 255),
    (80, 255, 220),
]





def _frame_idx_to_mmss(frame_idx: int, fps: float) -> str:
    sec = int(frame_idx / fps)
    mm = sec // 60
    ss = sec % 60
    return f"{mm:02d}:{ss:02d}"




# -------------------------
# Visualization utilities
# -------------------------
def _to_uint8_rgb(frame: np.ndarray) -> np.ndarray:
    if frame.dtype != np.uint8:
        frame = np.clip(frame * 255.0, 0, 255).astype(np.uint8)

    if frame.ndim == 2:
        frame = np.stack([frame] * 3, axis=-1)
    elif frame.ndim == 3 and frame.shape[-1] == 4:
        frame = frame[..., :3]
    return frame


def _load_font(font_size: int):
    try:
        return ImageFont.truetype("DejaVuSans.ttf", font_size)
    except Exception:
        return ImageFont.load_default()


def _build_closed_boundary_color_maps(
    ranges_closed: List[Tuple[int, int]],
    T: int,
    palette: List[Tuple[int, int, int]],
    end_exclusive = False,
):
    start_map, end_map = {}, {}
    for k, (s, e) in enumerate(ranges_closed):

        if end_exclusive:
            e = e - 1

        if s > e:
            s, e = e, s
        if e < 0 or s >= T:
            continue
        s = max(0, s)
        e = min(T - 1, e)
        rgba = (*palette[k % len(palette)], 255)
        start_map[s] = rgba
        end_map[e] = rgba
    return start_map, end_map




def visualize_concated_frames(
    frames: np.ndarray,
    out_dir: str,
    highlight_ranges_closed: Optional[List[Tuple[int, int]]],
    max_frames_per_img: int = 600,
    cols: int = 12,
    pad: int = 3,
    bg_color: Tuple[int, int, int] = (0, 0, 0),
    resize_to: Optional[Tuple[int, int]] = None,
    start_index: int = 0,           # time 的对齐的位置
    text_color: Tuple[int, int, int] = (255, 0, 0),
    font_size: int = 18,
    text_pad: Tuple[int, int] = (4, 2),
    draw_text_bg: bool = True,
    text_bg_rgba: Tuple[int, int, int, int] = (0, 0, 0, 160),
    out_prefix: str = "concat_",
    out_ext: str = ".jpg",
    jpg_quality: int = 75,
    bar_thickness: int = 14,
    palette: Optional[List[Tuple[int, int, int]]] = PALETTE,
    fps: Optional[float] = 24,
    end_range_exclusive = False,             # Whether end range is inclusive or exclusive
    verbose = False,
):

    os.makedirs(out_dir, exist_ok=True)
    font = _load_font(font_size)

    f0 = _to_uint8_rgb(frames[0])
    tile_h, tile_w = f0.shape[:2]
    frames_per_page = max_frames_per_img
    rows = math.ceil(frames_per_page / cols)

    canvas_w = cols * tile_w + (cols - 1) * pad
    canvas_h = rows * tile_h + (rows - 1) * pad

    def new_canvas():
        im = Image.new("RGB", (canvas_w, canvas_h), color=bg_color)
        return im, ImageDraw.Draw(im, "RGBA")

    start_map, end_map = ({}, {})
    if highlight_ranges_closed:
        start_map, end_map = _build_closed_boundary_color_maps(highlight_ranges_closed, len(frames), palette, end_exclusive=end_range_exclusive)

    page = 0
    global_idx = start_index
    canvas, draw = new_canvas()
    saved_paths = []

    # Iterate
    for i, fr in enumerate(frames):
        local_i = i % frames_per_page

        # Store
        if local_i == 0 and i > 0:
            saved_path = os.path.join(out_dir, f"{out_prefix}{page:04d}{out_ext}")
            canvas.save(saved_path, quality=jpg_quality)
            saved_paths.append(saved_path)

            # Update
            page += 1
            canvas, draw = new_canvas()
            

        r = local_i // cols
        c = local_i % cols
        x = c * (tile_w + pad)
        y = r * (tile_h + pad)

        pil = Image.fromarray(_to_uint8_rgb(fr))
        canvas.paste(pil, (x, y))

        if i in start_map:
            draw.rectangle((x, y, x + bar_thickness, y + tile_h), fill=start_map[i])
        if i in end_map:
            draw.rectangle((x + tile_w - bar_thickness, y, x + tile_w, y + tile_h), fill=end_map[i])

        text1 = str(global_idx)
        text2 = _frame_idx_to_mmss(global_idx, fps) if fps is not None else ""

        tx, ty = x + text_pad[0], y + text_pad[1]
        if draw_text_bg:
            bbox = draw.textbbox((tx, ty), text1 + "\n" + text2, font=font)
            draw.rectangle(bbox, fill=text_bg_rgba)

        draw.text((tx, ty), f"{text1}\n{text2}", font=font, fill=text_color)

        global_idx += 1


    # Save the last one
    saved_path = os.path.join(out_dir, f"{out_prefix}{page:04d}{out_ext}")
    canvas.save(saved_path, quality=jpg_quality)
    saved_paths.append(saved_path)

    
    # if verbose:
    #     print(f"Done. Total frames: {len(frames)} | Pages: {page + 1}")


    return saved_paths





def concat_image_lists_horizontal(
    list1: List[str],
    list2: List[str],
    out_dir: str,
    bar_width: int = 40,
    bar_color: Tuple[int, int, int] = (255, 0, 0),
    out_prefix: str = "merged_",
    out_ext: str = ".jpg",
    jpg_quality: int = 90,
    resize_mode: str = "match_height",  # ["match_height", "match_width", "none"]
    verbose: bool = True,
) -> List[str]:
    """
    Horizontally concatenate images from two path lists (same index),
    with a thick visual bar in between.

    Args:
        list1, list2: list of image file paths (must have same length)
        out_dir: directory to save merged images
        bar_width: thickness of the separator bar
        bar_color: RGB color of separator bar
        resize_mode:
            - "match_height": resize second image to match height
            - "match_width": resize second image to match width
            - "none": no resize (heights must match)
    Returns:
        List of saved image paths
    """

    assert len(list1) == len(list2), "Two lists must have same length"

    os.makedirs(out_dir, exist_ok=True)

    saved_paths = []

    for idx, (p1, p2) in enumerate(zip(list1, list2)):

        img1 = Image.open(p1).convert("RGB")
        img2 = Image.open(p2).convert("RGB")

        # ---------- Resize logic ----------
        if resize_mode == "match_height":
            if img1.height != img2.height:
                new_w = int(img2.width * img1.height / img2.height)
                img2 = img2.resize((new_w, img1.height), Image.BILINEAR)

        elif resize_mode == "match_width":
            if img1.width != img2.width:
                new_h = int(img2.height * img1.width / img2.width)
                img2 = img2.resize((img1.width, new_h), Image.BILINEAR)

        elif resize_mode == "none":
            assert img1.height == img2.height, \
                "Heights must match when resize_mode='none'"

        else:
            raise ValueError("resize_mode must be one of ['match_height', 'match_width', 'none']")

        # ---------- Create bar ----------
        bar = Image.new("RGB", (bar_width, img1.height), color=bar_color)

        # ---------- Create canvas ----------
        total_width = img1.width + bar_width + img2.width
        canvas = Image.new("RGB", (total_width, img1.height))

        canvas.paste(img1, (0, 0))
        canvas.paste(bar, (img1.width, 0))
        canvas.paste(img2, (img1.width + bar_width, 0))

        # ---------- Save ----------
        out_path = os.path.join(out_dir, f"{out_prefix}{idx:04d}{out_ext}")
        canvas.save(out_path, quality=jpg_quality)
        saved_paths.append(out_path)

    if verbose:
        print(f"Done. Saved {len(saved_paths)} merged images to {out_dir}")


    return saved_paths