AiDOPWarehouse/doc/_tmp_probe_offset.py

# -*- coding: utf-8 -*-
"""量出各页「单位名称文字中心」与「已盖红章中心」的偏差，得出校正量。"""
import io
import os
import sys

import fitz
import numpy as np

sys.stdout = io.TextIOWrapper(sys.stdout.buffer, encoding="utf-8")

HERE = os.path.dirname(os.path.abspath(__file__))
PDF = os.path.join(HERE, "_tmp_inspect", "sealed.pdf")
COMPANY = "北京智造易科技有限公司"
PAGES = [1, 5, 6, 7, 9, 12, 46, 61, 62, 65, 76, 79, 91]

d = fitz.open(PDF)
p0 = d[0]
print(f"page.rect {p0.rect}　mediabox {p0.mediabox}　rotation {p0.rotation}")
print(f"即 {p0.rect.width / 72 * 25.4:.1f}×{p0.rect.height / 72 * 25.4:.1f} mm\n")

print(f"{'页':>4} {'文字中心 x,y':>18} {'红章中心 x,y':>18} "
      f"{'Δx':>7} {'Δy':>7} {'章直径mm':>9}")
for pno in PAGES:
    pg = d[pno - 1]
    hits = pg.search_for(COMPANY)
    # 红章：取红色像素连通范围（整页扫描件页会误判，这些页没有扫描件）
    pm = pg.get_pixmap(dpi=100)
    a = np.frombuffer(pm.samples, dtype=np.uint8).reshape(pm.height, pm.width, pm.n)
    r, g, b = a[..., 0].astype(int), a[..., 1].astype(int), a[..., 2].astype(int)
    red = (r > 110) & (r - g > 55) & (r - b > 55)
    ys, xs = np.nonzero(red)
    if not len(ys) or not hits:
        print(f"{pno:>4}  未找到（文字 {len(hits)} 处，红像素 {len(ys)}）")
        continue
    sc = 72 / 100
    cx = (xs.min() + xs.max()) / 2 * sc
    cy = (ys.min() + ys.max()) / 2 * sc
    dia = (xs.max() - xs.min() + 1) * sc / 72 * 25.4
    # 文字命中可能多处，取与红章最近的
    best = min(hits, key=lambda rc: abs((rc.y0 + rc.y1) / 2 - cy))
    tx, ty = (best.x0 + best.x1) / 2, (best.y0 + best.y1) / 2
    print(f"{pno:>4} {tx:8.1f},{ty:8.1f} {cx:8.1f},{cy:8.1f} "
          f"{cx - tx:+7.1f} {cy - ty:+7.1f} {dia:9.1f}")
d.close()