coco_to_labelImg模块只存在bbox时，会出现丢失问题

[SWHL]

问题描述

当前代码假定COCO数据集中必然存在segmentation字段，才会正常转换。

LabelConvert/label_convert/coco_to_labelImg.py

Lines 73 to 83 in 620d19f

	seg_info = one_anno.get("segmentation")
	if seg_info:
	bbox = self.get_bbox(seg_info)
	xywh = self.xyxy_to_xywh(bbox, img_width, img_height)
	category_id = int(one_anno.get("category_id")) - 1
	xywh_str = " ".join([str(v) for v in xywh])
	label_str = f"{category_id} {xywh_str}"
	# 写入标注的txt文件
	txt_full_path = save_dir / f"{Path(img_name).stem}.txt"
	self.write_txt(txt_full_path, label_str, mode="a")

如果仅仅存在bbox字段，但是没有segmentation字段，则程序会为空！

所用版本

label_convert==0.1.2

[SWHL]

已经在label_convert>=0.1.3中修复