get_max_min_coordinates

本文介绍如何使用PCL库加载.pcd文件格式的点云数据,并通过pcl::getMinMax3D函数获取点云数据的最大最小边界坐标。代码示例展示了从指定路径加载点云,以及打印点云的边界框坐标。

摘要生成于 C知道 ,由 DeepSeek-R1 满血版支持, 前往体验 >

	pcl::PointCloud<pcl::PointXYZ>::Ptr cloud;
	cloud = pcl::PointCloud<pcl::PointXYZ>::Ptr(new pcl::PointCloud<pcl::PointXYZ>);
	std::string szPath = "C:\\Users\\BZL\\Desktop\\rs1.pcd";
	pcl::io::loadPCDFile<pcl::PointXYZ>(szPath, *cloud);
	pcl::PointXYZ minPt, maxPt;
	pcl::getMinMax3D(*cloud, minPt, maxPt);
	std::cout << "Max x: " << maxPt.x << std::endl;
	std::cout << "Max y: " << maxPt.y << std::endl;
	std::cout << "Max z: " << maxPt.z << std::endl;
	std::cout << "Min x: " << minPt.x << std::endl;
	std::cout << "Min y: " << minPt.y << std::endl;
	std::cout << "Min z: " << minPt.z << std::endl;

 

from paddleocr import PaddleOCR, draw_ocr import json from shapely.geometry import Polygon import os import time 初始化OCR模型 ocr = PaddleOCR(use_angle_cls=True, lang=“ch”) test_img_path = r"C:\Users\86178\Desktop\2\d5f5faec2b7f05652e82cb87ba85210.jpg" test_json_path = “f_template.json” with open(test_json_path, “r”, encoding=‘utf-8’) as f: data = json.load(f) print(data) === 步骤2:提取待匹配文本 === discretionary = [item.get(“key”, “”) for item in data[“data”]]#discretionary:目标key,字典 print(discretionary) offsets = [item[“offset”] for item in data[“data”]]#offsets:目标key的偏移量 print(offsets) 获取OCR识别结果 result = ocr.ocr(test_img_path, cls=True)#result:ocr结果 print(result) ocr_results = []#ocr_results:[{‘text’: ‘ASG’, ‘coordinates’: [[816.0, 143.0], [925.0, 138.0], [928.0, 200.0], [819.0, 205.0]]},这种格式的ocr结果 for line in result[0]: # result[0] 包含OCR识别的每行结果 key = line[1][0] # 假设识别到的文本是字段名 key_offset = line[0] # 获取字段名的坐标(边框) print(key) 存储OCR结果 ocr_results.append({ “text”: key, “coordinates”: key_offset }) print(ocr_results) ocr_dict = {item[“text”]: item[“coordinates”] for item in ocr_results}#ocr_dict:保存ocr结果中文本和坐标的数组 result_coordinates = []#result_coordinates:字典中的key在ocr结果中对应的坐标,没找到是[[0,0],[0,0],[0,0],[0,0]] for target_text in discretionary: result_coordinates.append( ocr_dict[target_text] if target_text in ocr_dict else [[0, 0] for _ in range(4)] ) print(result_coordinates) 处理逻辑 v_coordinates = []#v_coordinates:通过模板计算的value的大致坐标 for sub_arr, offset in zip(result_coordinates, offsets): new_sub = [] for point in sub_arr: new_x = point[0] - offset[‘x_offset’] new_y = point[1] - offset[‘y_offset’] new_sub.append([new_x, new_y]) v_coordinates.append(new_sub) print(v_coordinates) 将数组中的坐标转换为多边形列表 array_polygons = [Polygon(coords) for coords in v_coordinates] 将OCR结果中的坐标转换为(条目, 多边形)列表 ocr_entries = [ (entry, Polygon(entry[‘coordinates’])) for entry in ocr_results ] value = [] for array_poly in array_polygons: max_area = 0 best_match = None for entry, ocr_poly in ocr_entries: 计算交集面积 intersection = array_poly.intersection(ocr_poly) area = intersection.area 更新最大交集记录 if area > max_area: max_area = area best_match = entry 保存结果(若无交集,best_match为None) value.append(best_match) print(value) 遍历替换操作 修改遍历对象为 data[‘data’] for i in range(len(data[‘data’])): # 注意这里的 data[‘data’] data[‘data’][i][‘key_coordinates’] = result_coordinates[i] print(data) for data_item, value_item in zip(data[‘data’], value): if value_item is None: data_item[‘value’] = ‘’ # 或设置为None根据业务需求 data_item[‘value_coordinates’] = [] # 或设置为None else: data_item[‘value’] = value_item[‘text’] data_item[‘value_coordinates’] = value_item[‘coordinates’] print(data) for item in data[‘data’]: item[‘afterCorrection’] = ‘’ # 或设置为None def calculate_direction(key_coords, value_coords): 空值处理 if not value_coords: return None 提取四边形边界 def get_bounds(coords): xs = [p[0] for p in coords] ys = [p[1] for p in coords] return (min(xs), max(xs), min(ys), max(ys)) k_min_x, k_max_x, k_min_y, k_max_y = get_bounds(key_coords) v_min_x, v_max_x, v_min_y, v_max_y = get_bounds(value_coords) 方向判断优先级:右/左 → 下/上 → 中心点偏移 if v_min_x > k_max_x: # value整体在key右侧 return “右” elif v_max_x < k_min_x: # value整体在key左侧 return “左” elif v_min_y > k_max_y: # value整体在key下方(屏幕坐标系) return “下” elif v_max_y < k_min_y: # value整体在key上方 return “上” else: # 有重叠时,回退到中心点偏移逻辑 计算中心点 k_center_x = (k_min_x + k_max_x) / 2 k_center_y = (k_min_y + k_max_y) / 2 v_center_x = (v_min_x + v_max_x) / 2 v_center_y = (v_min_y + v_max_y) / 2 dx = v_center_x - k_center_x dy = v_center_y - k_center_y if abs(dx) > abs(dy): return “右” if dx > 0 else “左” else: return “下” if dy > 0 else “上” # 根据屏幕坐标系 更新direction字段 for item in data[‘data’]: key_coords = item[‘key_coordinates’] value_coords = item[‘value_coordinates’] item[‘direction’] = calculate_direction(key_coords, value_coords)将这个代码改成一个函数,可以在别的文件里面调用,输入是图片地址和模板地址,输出是modified_data
03-12
评论
添加红包

请填写红包祝福语或标题

红包个数最小为10个

红包金额最低5元

当前余额3.43前往充值 >
需支付:10.00
成就一亿技术人!
领取后你会自动成为博主和红包主的粉丝 规则
hope_wisdom
发出的红包
实付
使用余额支付
点击重新获取
扫码支付
钱包余额 0

抵扣说明:

1.余额是钱包充值的虚拟货币,按照1:1的比例进行支付金额的抵扣。
2.余额无法直接购买下载,可以购买VIP、付费专栏及课程。

余额充值