IBEM公式检测数据集标注使用方法

原创已于 2025-02-25 14:21:35 修改 · 427 阅读

2 ·

CC 4.0 BY-SA版权

文章标签：

#算法

于 2025-02-25 14:17:16 首次发布

IBEM Mathematical Formula Detection Dataset

label为如下格式

# ====================================
# class => {0: embedded, 1: isolated}
# ====================================
# x_rel   y_rel   width   height   class
28.75   28.56   42.29   3.08   1
36.07   38.38   27.37   3.08   1
37.32   46.58   25.16   3.08   1
74.64   15.82   4.77   1.07   0

展示矩形框

img = cv2.imread(img_path)
img_h, img_w = img.shape[:2]
x_rel, y_rel, w, h, cls = map(float, line.split())

# 将相对坐标转换为实际坐标
x1 = int(x_rel * img_w / 100.0)
y1 = int(y_rel * img_h / 100.0)
x2 = int((x_rel + w) * img_w / 100.0)
y2 = int((y_rel + h) * img_h / 100.0)

print(f"x1: {x1}, y1: {y1}, x2: {x2}, y2: {y2}")

# 边界框与标签
color = class_colors.get(0, (0, 255, 0))
cv2.rectangle(img, (x1, y1), (x2, y2), (0,255,0), 2)

pil = Image.fromarray(img)
pil.show()
cv2.imshow('img', img)
cv2.waitKey(0)