import math from pathlib import Path import cv2 import numpy as np import os import time def fry_cv2_imread(filename, flags=cv2.IMREAD_COLOR): try: with open(filename, 'rb') as f: chunk = f.read() chunk_arr = np.frombuffer(chunk, dtype=np.uint8) img = cv2.imdecode(chunk_arr, flags) if img is None: error_info = f"Warning: Unable to decode image: {filename}" print("警告", error_info) return img except IOError as e: error_info = f"IOError: Unable to read file: {filename}" print("错误", error_info) print("错误", f"Error details: {str(e)}") return None def fry_cv2_imwrite(filename, img, params=None): try: ext = os.path.splitext(filename)[1].lower() result, encoded_img = cv2.imencode(ext, img, params) if result: with open(filename, 'wb') as f: encoded_img.tofile(f) return True else: print("警告", f"Warning: Unable to encode image: {filename}") return False except Exception as e: print("错误", f"Error: Unable to write file: {filename}") print("错误", f"Error details: {str(e)}") return False # 覆盖 OpenCV 的原始函数 cv2.imread = fry_cv2_imread cv2.imwrite = fry_cv2_imwrite from fry_project_classes.blend_type_mixin import BlendTypeMixin class ImageStitcherTemplateMatch(BlendTypeMixin): def __init__(self, estimate_overlap_pixels=800, center_ratio=0.8, stitch_type="vertical", blend_type='half_importance', blend_ratio: float = 0.3, debug=False, debug_dir='debug_output', light_uniformity_compensation_enabled=False, light_uniformity_compensation_width=15, debug_draw_line_enabled=False ): """ 初始化拼图器 参数: overlap_pixels: 重叠区域像素数，默认800像素（预估值） center_ratio: 中心区域比例，默认0.8 debug: 是否开启调试模式，默认False debug_dir: 调试图片保存目录，默认'debug_output' use_weight_blend: 是否使用加权融合，默认True """ print("警告", f"拼图方法：区域") self.estimate_overlap_pixels = estimate_overlap_pixels self.estimate_non_overlap_pixels = None self.center_ratio = center_ratio self.blend_ratio = blend_ratio self.blend_type = blend_type self.stitch_type = stitch_type self.debug_draw_line_enabled = debug_draw_line_enabled self.debug = debug self.init_debug = debug_dir self.light_uniformity_compensation_enabled = light_uniformity_compensation_enabled self.light_uniformity_compensation_width = light_uniformity_compensation_width if self.debug: self.debug_dir = f"{self.init_debug}_{self.stitch_type}_{self.blend_type}" os.makedirs(self.debug_dir, exist_ok=True) os.makedirs(self.debug_dir, exist_ok=True) # 模板匹配参数 self.best_y = -1 self.best_x = -1 self.best_score = -1 self.template_score = -1 def save_debug_image(self, img, name, normalize=False): """ 保存调试图片 参数: img: 要保存的图片 name: 图片名称 normalize: 是否需要归一化处理（对于模板匹配结果图等） """ try: if self.debug: save_path = os.path.join(self.debug_dir, f"{name}.jpg") if normalize: # 归一化到0-255范围 img_normalized = cv2.normalize(img, None, 0, 255, cv2.NORM_MINMAX) cv2.imwrite(save_path, img_normalized) else: cv2.imwrite(save_path, img) now_info = f"Debug: Saved {save_path}" # print("信息", now_info) return True, f"save_debug_image 成功: {save_path}" else: return False, "debug mode is not enabled" except Exception as e: error_info = f"save_debug_image出现bug: {str(e)}" print("警告", error_info) return False, error_info def visualize_template_match(self, template, search_region, best_y, best_x): """ 可视化模板匹配结果 参数: template: 模板图片 search_region: 搜索区域 best_y, best_x: 最佳匹配位置 """ try: if self.debug: vis_img = search_region.copy() h, w = template.shape[:2] cv2.rectangle(vis_img, (best_x, best_y), (best_x + w, best_y + h), (0, 255, 0), 2) is_ok, msg = self.save_debug_image(vis_img, 'tp_140_template_match_visualization') return is_ok, msg else: return False, "debug mode is not enabled" except Exception as e: error_info = f"visualize_template_match 出现bug: {str(e)}" print("警告", error_info) return False, msg def split_image(self, img, is_left_top=True): """ 分割图片为重叠区域和非重叠区域 """ height, width = img.shape[:2] overlap_width = min(self.estimate_overlap_pixels, width // 2) non_overlap_width = width - overlap_width if is_left_top: non_overlap_region = img[:, :non_overlap_width] overlap_region = img[:, non_overlap_width:] if self.debug: self.save_debug_image(img, 'split_220_left_top_original') self.save_debug_image(non_overlap_region, 'split_240_left_top_non_overlap') self.save_debug_image(overlap_region, 'split_260_left_top_overlap') else: overlap_region = img[:, :overlap_width] non_overlap_region = img[:, overlap_width:] if self.debug: self.save_debug_image(img, 'split_320_right_bottom_original') self.save_debug_image(overlap_region, 'split_340_right_bottom_overlap') self.save_debug_image(non_overlap_region, 'split_360_right_bottom_non_overlap') return overlap_region, non_overlap_region def get_center_region(self, img): """ 获取图片的中心区域 """ height, width = img.shape[:2] margin_y = int(height * (1 - self.center_ratio) / 2) margin_x = int(width * (1 - self.center_ratio) / 2) center_region = img[margin_y:height - margin_y, margin_x:width - margin_x] if self.debug: self.save_debug_image(center_region, 'center_120_template_center_region') return center_region, margin_x, margin_y def template_matching(self, template, search_region): """ 模板匹配 """ result = cv2.matchTemplate(search_region, template, cv2.TM_CCOEFF_NORMED) min_val, max_val, min_loc, max_loc = cv2.minMaxLoc(result) best_y, best_x = max_loc[1], max_loc[0] now_info = f"匹配结果: 最大值{max_val}, 位置{best_x, best_y}" print("警告", now_info) self.best_y = best_y self.best_x = best_x self.best_score = max_val self.template_score = max_val # 如果存在父对象且父对象有 update_match_score 方法，发送匹配分数 # if hasattr(self, 'parent') and hasattr(self.parent, 'update_match_score'): # self.parent.update_match_score(max_val) if self.debug: is_ok1, opt_msg1 = self.save_debug_image(result, 'tp_120_template_matching_result', normalize=True) is_ok2, opt_msg2 = self.visualize_template_match(template, search_region, best_y, best_x) is_ok3, opt_msg3 = self.save_debug_image(template, 'tp_220_template') is_ok4, opt_msg4 = self.save_debug_image(search_region, 'tp_240_search_region') with open(os.path.join(self.debug_dir, 'tp_320_matching_info.txt'), 'w') as f: f.write(f"Best match position: ({best_x}, {best_y})\n") f.write(f"Match score: {max_val}\n") return best_x, best_y, max_val def pad_image(self, img: np.ndarray, target_width: int = None, target_height: int = None) -> np.ndarray: """ 将图片填充到目标尺寸 参数: img: 输入图片 target_width: 目标宽度，如果为None则保持原宽度 target_height: 目标高度，如果为None则保持原高度 返回: 填充后的图片 """ if self.debug: self.save_debug_image(img, 'pad_120_before_padding') current_height, current_width = img.shape[:2] target_width = target_width if target_width is not None else current_width target_height = target_height if target_height is not None else current_height if current_width == target_width and current_height == target_height: return img # 创建黑色背景 padded_img = np.zeros((target_height, target_width, 3), dtype=np.uint8) # 将原图放在中心位置 y_offset = (target_height - current_height) // 2 x_offset = (target_width - current_width) // 2 padded_img[y_offset:y_offset + current_height, x_offset:x_offset + current_width] = img if self.debug: self.save_debug_image(padded_img, 'pad_220_after_padding') with open(os.path.join(self.debug_dir, 'pad_320_padding_info.txt'), 'w') as f: f.write(f"Original size: {current_width}x{current_height}\n") f.write(f"Target size: {target_width}x{target_height}\n") f.write(f"Padding offsets: x={x_offset}, y={y_offset}\n") return padded_img def stitch_horizontal(self, left_img, right_img): """ 水平拼接两张图片 水平拼接的图片的高必须一样 """ # self.debug_dir = f"{self.init_debug}_horizontal_{self.blend_type}" # os.makedirs(self.debug_dir, exist_ok=True) # 确保两张图片高度相同 max_height = max(left_img.shape[0], right_img.shape[0]) left_img = self.pad_image(left_img, target_height=max_height) right_img = self.pad_image(right_img, target_height=max_height) # 1. 分割图片，使用预估的重叠像素 left_overlap, left_non_overlap = self.split_image(left_img, is_left_top=True) right_overlap, right_non_overlap = self.split_image(right_img, is_left_top=False) # 2. 获取左图重叠区域的中心部分作为模板 template, template_offset_x, template_offset_y = self.get_center_region(left_overlap) # 计算模板在左图中的位置 self.estimate_non_overlap_pixels = left_img.shape[1] - self.estimate_overlap_pixels template_in_left_x = self.estimate_non_overlap_pixels + template_offset_x template_in_left_y = template_offset_y # 3. 在右图重叠区域中进行模板匹配 best_x, best_y, max_val = self.template_matching(template, right_overlap) template_score = max_val # 计算模板在右图中的位置 template_in_right_x = best_x template_in_right_y = best_y # 真实的重叠区域的x和y 这个还不好算 # 04、计算最后拼接的图片的尺寸 left_width_contribution = template_in_left_x # 左图取到模板的左上角，不包含模板 right_width_contribution = right_img.shape[1] - template_in_right_x # 右图取到模板的左上角，包含模板 stitch_img_width = left_width_contribution + right_width_contribution stitch_img_height = max(left_img.shape[0], right_img.shape[0]) # 计算右图相对于左图的y方向的偏移量 y_offset_right2left = template_in_left_y - template_in_right_y # 计算真正的重叠区域：右图的左边+模板宽度+左图的右边 real_overlap_width = template_in_right_x + template.shape[1] + ( left_img.shape[1] - template_in_left_x - template.shape[1]) if self.debug: with open(os.path.join(self.debug_dir, 'h_320_alignment_info.txt'), 'w') as f: f.write(f"template_in_left_x: {template_in_left_x}\n") f.write(f"template_in_left_y: {template_in_left_y}\n") f.write(f"template_in_right_x: {template_in_right_x}\n") f.write(f"template_in_right_y: {template_in_right_y}\n") f.write(f"left_width_contribution: {left_width_contribution}\n") f.write(f"right_width_contribution: {right_width_contribution}\n") f.write(f"stitch_img_width: {stitch_img_width}\n") f.write(f"stitch_img_height: {stitch_img_height}\n") f.write(f"real_overlap_width: {real_overlap_width}\n") if self.blend_type == 'half_importance': blend_stitch_img = self.blend_half_importance(left_img, right_img, stitch_img_width, stitch_img_height, y_offset_right2left, real_overlap_width, light_uniformity_compensation=self.light_uniformity_compensation_enabled, light_uniformity_compensation_width=self.light_uniformity_compensation_width) if self.debug_draw_line_enabled: blend_stitch_img_visualize = self.blend_half_importance(left_img, right_img, stitch_img_width, stitch_img_height, y_offset_right2left, real_overlap_width, visualize=True) if self.debug: self.save_debug_image(blend_stitch_img_visualize, 'h_500_final_result_horizontal_visualize') elif self.blend_type == 'right_first': # 右边优先的拼接方式 blend_stitch_img = self.blend_right_first(left_img, right_img, stitch_img_width, stitch_img_height, y_offset_right2left) elif self.blend_type == 'left_first': blend_stitch_img = self.blend_left_first(left_img, right_img, stitch_img_width, stitch_img_height, y_offset_right2left, real_overlap_width) elif self.blend_type == 'half_importance_add_weight': blend_stitch_img = self.blend_half_importance_add_weight(left_img, right_img, stitch_img_width, stitch_img_height, y_offset_right2left, real_overlap_width, blend_ratio=self.blend_ratio) elif self.blend_type == 'half_importance_global_brightness': blend_stitch_img = self.blend_half_importance_global_brightness(left_img, right_img, stitch_img_width, stitch_img_height, y_offset_right2left, real_overlap_width, light_uniformity_compensation=self.light_uniformity_compensation_enabled, light_uniformity_compensation_width=self.light_uniformity_compensation_width) if self.debug_draw_line_enabled: blend_stitch_img_visualize = self.blend_half_importance_global_brightness(left_img, right_img, stitch_img_width, stitch_img_height, y_offset_right2left, real_overlap_width, visualize=True) if self.debug: self.save_debug_image(blend_stitch_img_visualize, 'h_500_final_result_horizontal_visualize') elif self.blend_type == 'half_importance_partial_brightness': blend_stitch_img = self.blend_half_importance_partial_brightness(left_img, right_img, stitch_img_width, stitch_img_height, y_offset_right2left, real_overlap_width, light_uniformity_compensation=self.light_uniformity_compensation_enabled, light_uniformity_compensation_width=self.light_uniformity_compensation_width) if self.debug_draw_line_enabled: blend_stitch_img_visualize = self.blend_half_importance_partial_brightness(left_img, right_img, stitch_img_width, stitch_img_height, y_offset_right2left, real_overlap_width, visualize=True) if self.debug: self.save_debug_image(blend_stitch_img_visualize, 'h_500_final_result_horizontal_visualize') elif self.blend_type == 'blend_half_importance_partial_HV': blend_stitch_img = self.blend_half_importance_partial_HV(left_img, right_img, stitch_img_width, stitch_img_height, y_offset_right2left, real_overlap_width, light_uniformity_compensation=self.light_uniformity_compensation_enabled, light_uniformity_compensation_width=self.light_uniformity_compensation_width) if self.debug_draw_line_enabled: blend_stitch_img_visualize = self.blend_half_importance_partial_HV(left_img, right_img, stitch_img_width, stitch_img_height, y_offset_right2left, real_overlap_width, visualize=True) if self.debug: self.save_debug_image(blend_stitch_img_visualize, 'h_500_final_result_horizontal_visualize') elif self.blend_type == 'blend_half_importance_partial_SV': blend_stitch_img = self.blend_half_importance_partial_SV(left_img, right_img, stitch_img_width, stitch_img_height, y_offset_right2left, real_overlap_width, light_uniformity_compensation=self.light_uniformity_compensation_enabled, light_uniformity_compensation_width=self.light_uniformity_compensation_width) if self.debug_draw_line_enabled: blend_stitch_img_visualize = self.blend_half_importance_partial_SV(left_img, right_img, stitch_img_width, stitch_img_height, y_offset_right2left, real_overlap_width, visualize=True) if self.debug: self.save_debug_image(blend_stitch_img_visualize, 'h_500_final_result_horizontal_visualize') elif self.blend_type == 'blend_half_importance_partial_HSV': blend_stitch_img = self.blend_half_importance_partial_HSV(left_img, right_img, stitch_img_width, stitch_img_height, y_offset_right2left, real_overlap_width, light_uniformity_compensation=self.light_uniformity_compensation_enabled, light_uniformity_compensation_width=self.light_uniformity_compensation_width) if self.debug_draw_line_enabled: blend_stitch_img_visualize = self.blend_half_importance_partial_HSV(left_img, right_img, stitch_img_width, stitch_img_height, y_offset_right2left, real_overlap_width, visualize=True) if self.debug: self.save_debug_image(blend_stitch_img_visualize, 'h_500_final_result_horizontal_visualize') elif self.blend_type == 'blend_half_importance_partial_brightness_add_weight': blend_stitch_img = self.blend_half_importance_partial_brightness_add_weight( left_img=left_img, right_img=right_img, stitch_img_width=stitch_img_width, stitch_img_height=stitch_img_height, y_offset_right2left=y_offset_right2left, real_overlap_width=real_overlap_width, light_uniformity_compensation=self.light_uniformity_compensation_enabled, light_uniformity_compensation_width=self.light_uniformity_compensation_width, add_weight_rate=self.blend_ratio ) if self.debug_draw_line_enabled: blend_stitch_img_visualize = self.blend_half_importance_partial_brightness_add_weight(left_img=left_img, right_img=right_img, stitch_img_width=stitch_img_width, stitch_img_height=stitch_img_height, y_offset_right2left=y_offset_right2left, real_overlap_width=real_overlap_width, add_weight_rate=self.blend_ratio, visualize=True) if self.debug: self.save_debug_image(blend_stitch_img_visualize, 'h_500_final_result_horizontal_visualize') else: # 左边优先的拼接方式 blend_stitch_img = None if self.debug: self.save_debug_image(blend_stitch_img, 'h_500_final_result_horizontal') # return result return blend_stitch_img def stitch_vertical(self, top_img, bottom_img): """ 垂直拼接两张图片 垂直拼接的图片的宽必须一样 """ # self.debug_dir = f"{self.init_debug}_vertical_{self.blend_type}" # os.makedirs(self.debug_dir, exist_ok=True) # 确保两张图片宽度相同 max_width = max(top_img.shape[1], bottom_img.shape[1]) top_img = self.pad_image(top_img, target_width=max_width) bottom_img = self.pad_image(bottom_img, target_width=max_width) if self.debug: self.save_debug_image(top_img, 'v_120_top_original') self.save_debug_image(bottom_img, 'v_140_bottom_original') # 将图片逆时针旋转90度 top_rotated = cv2.rotate(top_img, cv2.ROTATE_90_COUNTERCLOCKWISE) bottom_rotated = cv2.rotate(bottom_img, cv2.ROTATE_90_COUNTERCLOCKWISE) if self.debug: self.save_debug_image(top_rotated, 'v_220_top_rotated') self.save_debug_image(bottom_rotated, 'v_240_bottom_rotated') # 进行水平拼接 result_rotated = self.stitch_horizontal(top_rotated, bottom_rotated) # 将结果顺时针旋转90度 result = cv2.rotate(result_rotated, cv2.ROTATE_90_CLOCKWISE) if self.debug: self.save_debug_image(result, 'v_500_final_result_vertical') return result def stitch_main(self, left_img, right_img): if self.stitch_type == 'horizontal': # 确保两张图片高度相同 max_height = max(left_img.shape[0], right_img.shape[0]) left_img = self.pad_image(left_img, target_height=max_height) right_img = self.pad_image(right_img, target_height=max_height) return self.stitch_horizontal(left_img, right_img) elif self.stitch_type == 'vertical': # 确保两张图片宽度相同 max_width = max(left_img.shape[1], right_img.shape[1]) left_img = self.pad_image(left_img, target_width=max_width) right_img = self.pad_image(right_img, target_width=max_width) return self.stitch_vertical(left_img, right_img) else: raise ValueError('Invalid stitch type, must be one of horizontal or vertical') def main(): timestamp = time.strftime("%Y%m%d_%H%M%S") root_path = r"D:\_241231_fry_gitlab\LA_ai_main_CV_OpenCV\740_project\_250115_Stitch_Image_TemplateMatch\test_images" root_path_obj = Path(root_path).absolute() stitch_type = "vertical" debug_dir_str = str(root_path_obj / f'debug_{timestamp}_{stitch_type}') debug_dir_obj = Path(debug_dir_str).absolute() estimate_overlap_ratio = 0.45 estimate_overlap_pixels = int(round(1024 * estimate_overlap_ratio)) # 使用加权融合 # stitcher_weight = ImageStitcher( # estimate_overlap_pixels=estimate_overlap_pixels, # center_ratio=0.8, # debug=True, # debug_dir=debug_dir+"_weight", # use_weight_blend=True # ) # 读取图片 left_img_name = r"20250123_162407_0001.jpg" right_img_name = r"20250123_162409_0002.jpg" bottom_img_name = r"20250123_162422_0007.jpg" left_img_path = str(root_path_obj / left_img_name) right_img_path = str(root_path_obj / right_img_name) bottom_img_path = str(root_path_obj / bottom_img_name) left_img = cv2.imread(left_img_path) right_img = cv2.imread(right_img_path) top_img = left_img bottom_img = cv2.imread(bottom_img_path) start_time = time.time() # 使用简单拼接 stitcher_simple = ImageStitcherTemplateMatch( estimate_overlap_pixels=estimate_overlap_pixels, center_ratio=0.8, blend_type='half_importance_add_weight', # blend_type='left_first', stitch_type=stitch_type, blend_ratio=0.5, debug=True, debug_dir=debug_dir_str, ) # 竖直拼接 - 简单拼接 result_img = stitcher_simple.stitch_main(left_img, bottom_img) save_final_image_path = str(debug_dir_obj / 'result_img.jpg') cv2.imwrite(save_final_image_path, result_img) end_time = time.time() print(f"拼接图片耗时：{end_time - start_time:.2f}秒") if __name__ == '__main__': main()