#!/usr/bin/env python # -*- coding: utf-8 -*- """ 视频封面生成器 支持从视频提取帧并添加标题文字 """ import argparse import json import os import sys from PIL import Image, ImageDraw, ImageFont, ImageFilter import cv2 # ⚠️ 修复模块路径：确保 Python 能找到 modules 和其他本地模块 # 将脚本所在目录（python 目录）添加到 sys.path script_dir = os.path.dirname(os.path.abspath(__file__)) if script_dir not in sys.path: sys.path.insert(0, script_dir) print(f"[PATH] Added to sys.path: {script_dir}", file=sys.stderr) def extract_frame_from_video(video_path, timestamp=0): """从视频中提取指定时间的帧""" try: cap = cv2.VideoCapture(video_path) # 设置到指定时间（秒） cap.set(cv2.CAP_PROP_POS_MSEC, timestamp * 1000) ret, frame = cap.read() cap.release() if not ret: return None # 转换BGR到RGB frame_rgb = cv2.cvtColor(frame, cv2.COLOR_BGR2RGB) return Image.fromarray(frame_rgb) except Exception as e: print(f"Error extracting frame: {e}", file=sys.stderr) return None def hex_to_rgb(hex_color): """将十六进制颜色转换为RGB元组""" if isinstance(hex_color, str): hex_color = hex_color.lstrip('#') return tuple(int(hex_color[i:i+2], 16) for i in (0, 2, 4)) return tuple(hex_color) def add_text_to_image(img, title, config): """在图像上添加标题文字""" pil_img = img.copy().convert('RGBA') width, height = pil_img.size # 获取配置参数 - 支持新格式（来自前端的配置） # 新格式：titleFontSize, titleColor, titleFontFamily, titlePosition, titleStrokeWidth, titleStrokeColor # 旧格式：fontSize, fontColor, fontPath, position, style font_size = config.get('titleFontSize', config.get('fontSize', int(height * 0.08))) # 处理颜色格式：新格式是十六进制字符串（如 "#eb1414"），旧格式是RGB元组 title_color = config.get('titleColor', config.get('fontColor', '#FFFFFF')) font_color = hex_to_rgb(title_color) # 阴影参数 shadow_color = hex_to_rgb(config.get('titleShadowColor', '#000000')) shadow_offset_x = config.get('titleShadowOffsetX', 0) shadow_offset_y = config.get('titleShadowOffsetY', 0) shadow_blur_raw = config.get('titleShadowBlur', 0) # 修复：PIL的GaussianBlur效果比CSS重，需要缩小系数以匹配前端预览 # 经验值：CSS box-shadow blur 与 PIL GaussianBlur 的比例约为 2.5:1 shadow_blur = int(shadow_blur_raw * 0.4) if shadow_blur_raw > 0 else 0 # 描边参数 stroke_width = config.get('titleStrokeWidth', 0) stroke_color = hex_to_rgb(config.get('titleStrokeColor', '#000000')) print(f"[DEBUG] Text rendering params: stroke_width={stroke_width}, stroke_color={stroke_color}", file=sys.stderr) font_family = config.get('titleFontFamily', config.get('fontPath', 'SimHei')) # 使用 font_manager 查找字体（支持系统字体和ziti目录字体） font_path = None try: from modules.font_manager import get_font_manager font_manager = get_font_manager() font_path = font_manager.find_font(font_family, 400) if font_path: print(f"[DEBUG] Font found via font_manager: {font_family} -> {font_path}", file=sys.stderr) else: print(f"[DEBUG] Font not found via font_manager: {font_family}, using fallback", file=sys.stderr) except Exception as e: print(f"[DEBUG] font_manager not available: {e}, using hardcoded fonts", file=sys.stderr) # 回退到硬编码字体路径 if not font_path: if font_family == 'SimHei': font_path = 'C:\\Windows\\Fonts\\simhei.ttf' elif font_family == 'SimSun': font_path = 'C:\\Windows\\Fonts\\simsun.ttc' elif font_family == 'Microsoft YaHei': font_path = 'C:\\Windows\\Fonts\\msyh.ttc' else: font_path = config.get('fontPath', 'C:\\Windows\\Fonts\\msyh.ttc') print(f"[DEBUG] Using hardcoded font path: {font_path}", file=sys.stderr) # 位置：新格式是百分比坐标 {x, y}，旧格式是 'top', 'center', 'bottom' title_position = config.get('titlePosition', {}) if isinstance(title_position, dict): position = 'custom' else: position = config.get('position', 'top') # top, center, bottom style = config.get('style', 'default') # default, outline, blur_bg, gradient, split # 加载字体 try: font = ImageFont.truetype(font_path, font_size) except: font = ImageFont.load_default() # 创建绘图对象 draw = ImageDraw.Draw(pil_img) # 计算文字尺寸 bbox = draw.textbbox((0, 0), title, font=font) text_width = bbox[2] - bbox[0] text_height = bbox[3] - bbox[1] # 计算文字位置 if position == 'custom' and isinstance(title_position, dict): # 使用自定义位置（百分比） x = int(width * title_position.get('x', 50) / 100) - text_width // 2 y = int(height * title_position.get('y', 80) / 100) - text_height // 2 else: x = (width - text_width) // 2 if position == 'top': y = int(height * 0.08) elif position == 'center': y = (height - text_height) // 2 elif position == 'bottom': y = int(height * 0.85) - text_height else: y = int(height * 0.08) # 辅助函数：绘制带阴影和描边的文字 def draw_text_with_effects(draw_obj, pos_x, pos_y, text, font, color, stroke_w, stroke_c, shadow_ox, shadow_oy, shadow_b, shadow_c): """绘制带阴影和描边的文字""" # 1. 先绘制阴影（如果有偏移） if shadow_ox != 0 or shadow_oy != 0: if shadow_b > 0: # 创建阴影图层并模糊 shadow_layer = Image.new('RGBA', pil_img.size, (0, 0, 0, 0)) shadow_draw = ImageDraw.Draw(shadow_layer) shadow_draw.text((pos_x + shadow_ox, pos_y + shadow_oy), text, font=font, fill=shadow_c + (180,)) shadow_layer = shadow_layer.filter(ImageFilter.GaussianBlur(radius=shadow_b)) return shadow_layer else: draw_obj.text((pos_x + shadow_ox, pos_y + shadow_oy), text, font=font, fill=shadow_c + (180,)) # 2. 绘制描边 if stroke_w > 0: for adj_x in range(-stroke_w, stroke_w + 1): for adj_y in range(-stroke_w, stroke_w + 1): if adj_x != 0 or adj_y != 0: draw_obj.text((pos_x + adj_x, pos_y + adj_y), text, font=font, fill=stroke_c + (255,)) # 3. 绘制主文字 draw_obj.text((pos_x, pos_y), text, font=font, fill=color + (255,)) return None # 应用样式 if style == 'default': # 默认风格：黑色半透明背景条 padding = 30 overlay = Image.new('RGBA', pil_img.size, (0, 0, 0, 0)) overlay_draw = ImageDraw.Draw(overlay) overlay_draw.rectangle( [x - padding, y - padding, x + text_width + padding, y + text_height + padding], fill=(0, 0, 0, 180) ) pil_img = Image.alpha_composite(pil_img, overlay) draw = ImageDraw.Draw(pil_img) # 绘制阴影 if shadow_offset_x != 0 or shadow_offset_y != 0: if shadow_blur > 0: shadow_layer = Image.new('RGBA', pil_img.size, (0, 0, 0, 0)) shadow_draw = ImageDraw.Draw(shadow_layer) shadow_draw.text((x + shadow_offset_x, y + shadow_offset_y), title, font=font, fill=shadow_color + (180,)) shadow_layer = shadow_layer.filter(ImageFilter.GaussianBlur(radius=shadow_blur)) pil_img = Image.alpha_composite(pil_img, shadow_layer) draw = ImageDraw.Draw(pil_img) else: draw.text((x + shadow_offset_x, y + shadow_offset_y), title, font=font, fill=shadow_color + (180,)) # 绘制描边 if stroke_width > 0: print(f"[DEBUG] Drawing text stroke: width={stroke_width}, color={stroke_color}", file=sys.stderr) for adj_x in range(-stroke_width, stroke_width + 1): for adj_y in range(-stroke_width, stroke_width + 1): if adj_x != 0 or adj_y != 0: draw.text((x + adj_x, y + adj_y), title, font=font, fill=stroke_color + (255,)) else: print(f"[DEBUG] Stroke disabled: stroke_width={stroke_width}", file=sys.stderr) draw.text((x, y), title, font=font, fill=font_color + (255,)) elif style == 'outline': # 描边风格 # 绘制阴影 if shadow_offset_x != 0 or shadow_offset_y != 0: if shadow_blur > 0: shadow_layer = Image.new('RGBA', pil_img.size, (0, 0, 0, 0)) shadow_draw = ImageDraw.Draw(shadow_layer) shadow_draw.text((x + shadow_offset_x, y + shadow_offset_y), title, font=font, fill=shadow_color + (180,)) shadow_layer = shadow_layer.filter(ImageFilter.GaussianBlur(radius=shadow_blur)) pil_img = Image.alpha_composite(pil_img, shadow_layer) draw = ImageDraw.Draw(pil_img) else: draw.text((x + shadow_offset_x, y + shadow_offset_y), title, font=font, fill=shadow_color + (180,)) outline_width = stroke_width if stroke_width > 0 else 3 outline_color = stroke_color for adj_x in range(-outline_width, outline_width + 1): for adj_y in range(-outline_width, outline_width + 1): draw.text((x + adj_x, y + adj_y), title, font=font, fill=outline_color + (255,)) draw.text((x, y), title, font=font, fill=font_color + (255,)) elif style == 'blur_bg': # 模糊背景风格 padding = 50 bg_region = pil_img.crop(( max(0, x - padding), max(0, y - padding), min(width, x + text_width + padding), min(height, y + text_height + padding) )) bg_region = bg_region.filter(ImageFilter.GaussianBlur(radius=15)) pil_img.paste(bg_region, (max(0, x - padding), max(0, y - padding))) # 添加半透明遮罩 overlay = Image.new('RGBA', pil_img.size, (0, 0, 0, 0)) overlay_draw = ImageDraw.Draw(overlay) overlay_draw.rectangle( [x - padding, y - padding, x + text_width + padding, y + text_height + padding], fill=(0, 0, 0, 120) ) pil_img = Image.alpha_composite(pil_img, overlay) draw = ImageDraw.Draw(pil_img) draw.text((x, y), title, font=font, fill=font_color + (255,)) elif style == 'gradient': # 渐变背景风格 gradient_height = text_height + 100 gradient = Image.new('RGBA', (width, gradient_height), (0, 0, 0, 0)) gradient_draw = ImageDraw.Draw(gradient) for i in range(gradient_height): alpha = int(200 * (i / gradient_height)) gradient_draw.rectangle([0, i, width, i + 1], fill=(0, 0, 0, alpha)) pil_img.paste(gradient, (0, max(0, y - 50)), gradient) draw = ImageDraw.Draw(pil_img) draw.text((x, y), title, font=font, fill=font_color + (255,)) elif style == 'split': # 分栏风格 padding = 40 left_x = padding # 左侧标题背景 overlay = Image.new('RGBA', pil_img.size, (0, 0, 0, 0)) overlay_draw = ImageDraw.Draw(overlay) overlay_draw.rectangle( [left_x - 20, y - 20, left_x + text_width + 20, y + text_height + 20], fill=(0, 0, 0, 180) ) pil_img = Image.alpha_composite(pil_img, overlay) draw = ImageDraw.Draw(pil_img) draw.text((left_x, y), title, font=font, fill=font_color + (255,)) # 右侧装饰 accent_text = config.get('accentText', '✨') accent_bbox = draw.textbbox((0, 0), accent_text, font=font) accent_width = accent_bbox[2] - accent_bbox[0] right_x = width - accent_width - padding overlay = Image.new('RGBA', pil_img.size, (0, 0, 0, 0)) overlay_draw = ImageDraw.Draw(overlay) overlay_draw.rectangle( [right_x - 25, y - 25, right_x + accent_width + 25, y + text_height + 25], fill=(255, 193, 7, 200) ) pil_img = Image.alpha_composite(pil_img, overlay) draw = ImageDraw.Draw(pil_img) draw.text((right_x, y), accent_text, font=font, fill=(0, 0, 0, 255)) return pil_img.convert('RGB') def main(): parser = argparse.ArgumentParser(description='Generate video cover with title') parser.add_argument('--video', required=True, help='Path to video file') parser.add_argument('--title', required=True, help='Title text') parser.add_argument('--output', required=True, help='Output image path') parser.add_argument('--config', default='{}', help='JSON configuration') parser.add_argument('--preview', action='store_true', help='Preview mode') parser.add_argument('--timestamp', type=float, default=0, help='Frame timestamp in seconds') args = parser.parse_args() try: # 解析配置 config = json.loads(args.config) # 诊断日志：输出接收到的配置参数 print(f"[DEBUG] Received config: {json.dumps(config, indent=2, ensure_ascii=False)}", file=sys.stderr) print(f"[DEBUG] titleStrokeWidth: {config.get('titleStrokeWidth', 'NOT SET')}", file=sys.stderr) print(f"[DEBUG] titleStrokeColor: {config.get('titleStrokeColor', 'NOT SET')}", file=sys.stderr) print(f"[DEBUG] titleFontFamily: {config.get('titleFontFamily', 'NOT SET')}", file=sys.stderr) # 检查输入是图像还是视频 video_path = args.video if video_path.lower().endswith(('.jpg', '.jpeg', '.png', '.bmp', '.webp')): # 直接加载图像 print(f"Loading image: {video_path}") try: frame = Image.open(video_path) except Exception as e: print(f"Error: Failed to load image: {e}", file=sys.stderr) sys.exit(1) else: # 提取视频帧 print(f"Extracting frame from video: {video_path}") frame = extract_frame_from_video(video_path, args.timestamp) if frame is None: print("Error: Failed to extract frame from video", file=sys.stderr) sys.exit(1) # 调整图片大小 width, height = frame.size max_width = config.get('maxWidth', 1920) if width > max_width: scale = max_width / width new_width = max_width new_height = int(height * scale) frame = frame.resize((new_width, new_height), Image.Resampling.LANCZOS) # 添加标题 print(f"Adding title: {args.title}") result = add_text_to_image(frame, args.title, config) # 保存结果 os.makedirs(os.path.dirname(args.output), exist_ok=True) result.save(args.output, quality=95) print(f"Cover saved to: {args.output}") sys.exit(0) except Exception as e: print(f"Error: {e}", file=sys.stderr) import traceback traceback.print_exc() sys.exit(1) if __name__ == '__main__': main()