feat: 视频渲染后随机边缘裁剪 2-5% 降重 #1664
CI/CD Pipeline / Check push changed paths (pull_request) Has been skipped
CI/CD Pipeline / Check if frontend-only change (pull_request) Successful in 6s
CI/CD Pipeline / Dedup Check - skip PR tests when covered by push pipeline (pull_request) Successful in 7s
CI/CD Pipeline / Build Staging API Image (pull_request) Has been skipped
CI/CD Pipeline / Build Staging Web Image (pull_request) Has been skipped
CI/CD Pipeline / Build Staging Worker Image (pull_request) Has been skipped
Preview Deploy / Deploy Preview Environment (pull_request) Successful in 2m10s
CI/CD Pipeline / Frontend Lint (pull_request) Has been skipped
CI/CD Pipeline / Frontend Unit Tests (pull_request) Has been skipped
CI/CD Pipeline / PR Build API Image (pull_request) Successful in 25s
CI/CD Pipeline / PR Build Web Image (pull_request) Has been skipped
CI/CD Pipeline / PR Build Worker Image (pull_request) Successful in 16s
CI/CD Pipeline / Retag skipped Staging API Image (pull_request) Has been skipped
CI/CD Pipeline / Retag skipped Staging Web Image (pull_request) Has been skipped
CI/CD Pipeline / Retag skipped Staging Worker Image (pull_request) Has been skipped
CI/CD Pipeline / Deploy Staging (Watchtower auto-deploy) (pull_request) Has been skipped
CI/CD Pipeline / Staging E2E Tests (pull_request) Has been skipped
CI/CD Pipeline / Staging API Integration Tests (pull_request) Has been skipped
CI/CD Pipeline / ACR Image Cleanup (pull_request) Has been skipped
CI/CD Pipeline / Validate - Style (pull_request) Has been cancelled
CI/CD Pipeline / Validate - Security (pull_request) Has been cancelled
CI/CD Pipeline / Validate - Python (mypy + alembic) (pull_request) Has been cancelled
CI/CD Pipeline / Unit Tests (pull_request) Has been cancelled
CI/CD Pipeline / Integration Tests (pull_request) Has been cancelled
CI/CD Pipeline / Build Production API Image (pull_request) Has been cancelled
CI/CD Pipeline / Build Production Web Image (pull_request) Has been cancelled
CI/CD Pipeline / Build Production Worker Image (pull_request) Has been cancelled
CI/CD Pipeline / Deploy Production (pull_request) Has been cancelled
CI/CD Pipeline / Production Browser E2E (pull_request) Has been cancelled
CI/CD Pipeline / Canary Release to Production (pull_request) Has been cancelled
CI/CD Pipeline / CI Gate (pull_request) Has been cancelled
AI Code Review / AI Code Review (pull_request) Has been cancelled
PR Automation / Auto Approve on CI Green (pull_request) Has been cancelled
PR Automation / Auto Merge on CI Green + Approved (pull_request) Has been cancelled

- ffmpeg_utils.py 新增 random_edge_crop() 函数:
  - ffprobe 获取原始分辨率
  - 四边各自 random.uniform(2%, 5%) 裁剪
  - crop + scale 滤镜保持输出尺寸不变
  - 裁剪后奇数尺寸自动调整为偶数(ffmpeg 编码器要求)
  - 失败时 warning 不中断,回退使用原始视频

- generation.py 渲染完成后、_upload_and_record 前插入裁剪步骤
  - try/except 包裹,失败仅 log warning
  - gen_task 日志记录裁剪结果

- 9 个单测覆盖:基本流程/滤镜参数/编码参数/错误处理/偶数尺寸
- black + ruff 通过
This commit is contained in:
saas-backend-agent
2026-09-04 11:42:22 +08:00
parent 0542654ca8
commit a4ef63d76c
3 changed files with 353 additions and 0 deletions
@@ -304,3 +304,127 @@ def normalize_video(
]
run_ffmpeg(command)
return {"width": width, "height": height, "path": output_path}
def random_edge_crop(
input_path: str | Path,
output_path: str | Path | None = None,
*,
min_crop_pct: float = 0.02,
max_crop_pct: float = 0.05,
) -> Path:
"""对视频四边做随机裁剪再缩放回原分辨率,用于改变 pHash 指纹。
Args:
input_path: 输入视频路径
output_path: 输出路径;为 None 时写入 input_path 同目录的临时文件,
成功后覆盖原文件
min_crop_pct: 每边最小裁剪比例(默认 2%
max_crop_pct: 每边最大裁剪比例(默认 5%
Returns:
输出文件路径(Path 对象)
Raises:
subprocess.CalledProcessError: ffmpeg 执行失败时抛出
"""
import random
import shutil
import tempfile
input_path = Path(input_path)
# 获取原始分辨率
info = probe_video_info(str(input_path))
W = info["width"]
H = info["height"]
if W <= 0 or H <= 0:
logger.warning("无法获取视频分辨率 (W=%d H=%d),跳过裁剪: %s", W, H, input_path)
return input_path
# 四边各自随机裁剪 2%~5%
crop_top = int(H * random.uniform(min_crop_pct, max_crop_pct))
crop_bottom = int(H * random.uniform(min_crop_pct, max_crop_pct))
crop_left = int(W * random.uniform(min_crop_pct, max_crop_pct))
crop_right = int(W * random.uniform(min_crop_pct, max_crop_pct))
# 裁剪后尺寸(确保至少 2 像素)
new_w = max(W - crop_left - crop_right, 2)
new_h = max(H - crop_top - crop_bottom, 2)
x_offset = crop_left
y_offset = crop_top
# 确保裁剪尺寸为偶数(ffmpeg 编码器常要求偶数尺寸)
new_w = new_w if new_w % 2 == 0 else new_w - 1
new_h = new_h if new_h % 2 == 0 else new_h - 1
if new_w < 2:
new_w = 2
if new_h < 2:
new_h = 2
# 输出分辨率必须与原始一致
out_w = W if W % 2 == 0 else W + 1
out_h = H if H % 2 == 0 else H + 1
vf = f"crop={new_w}:{new_h}:{x_offset}:{y_offset},scale={out_w}:{out_h}"
logger.info(
"随机边缘裁剪: %s → crop(%d,%d,%d,%d)=%dx%d scale→%dx%d",
input_path.name,
crop_top,
crop_bottom,
crop_left,
crop_right,
new_w,
new_h,
out_w,
out_h,
)
# 确定输出路径
if output_path is None:
temp_fd, temp_path = tempfile.mkstemp(suffix=".mp4", dir=input_path.parent)
import os
os.close(temp_fd)
temp_output = Path(temp_path)
replace_original = True
else:
temp_output = Path(output_path)
replace_original = False
command = [
FFMPEG_BIN,
"-y",
"-i",
str(input_path),
"-vf",
vf,
"-c:v",
"libx264",
"-preset",
"fast",
"-crf",
"18",
"-c:a",
"copy",
"-movflags",
"+faststart",
str(temp_output),
]
try:
run_ffmpeg(command)
except Exception:
# 裁剪失败时清理临时文件
if temp_output.exists() and replace_original:
temp_output.unlink(missing_ok=True)
raise
# 成功 → 覆盖原文件
if replace_original:
shutil.move(str(temp_output), str(input_path))
return input_path
return temp_output