你是不是也經(jīng)常遇到這樣的場(chǎng)景項(xiàng)目文檔需要統(tǒng)一調(diào)整圖片尺寸幾十張照片要批量壓縮上傳或者收到一堆掃描版PDF需要提取文字和圖片手動(dòng)一張張?zhí)幚聿粌H耗時(shí)費(fèi)力還容易出錯(cuò)。最近在整理技術(shù)文檔時(shí)我發(fā)現(xiàn)了一個(gè)高效的解決方案——通過(guò)Python腳本實(shí)現(xiàn)圖片和PDF的批量處理全流程。這套方法真正解決了重復(fù)性工作的痛點(diǎn)讓原本需要半天的手工操作在幾分鐘內(nèi)自動(dòng)完成。本文將分享從環(huán)境搭建到完整實(shí)戰(zhàn)的詳細(xì)流程包含具體的代碼實(shí)現(xiàn)和常見問(wèn)題排查。無(wú)論你是需要處理技術(shù)文檔、項(xiàng)目資料還是日常辦公文件這套方案都能顯著提升效率。1. 核心痛點(diǎn)與解決方案對(duì)比在技術(shù)文檔管理、項(xiàng)目資料整理等場(chǎng)景中我們經(jīng)常面臨以下典型問(wèn)題傳統(tǒng)手工處理的痛點(diǎn)圖片尺寸不統(tǒng)一需要逐張調(diào)整寬高比大量圖片需要壓縮以減少存儲(chǔ)空間PDF文檔中的圖片需要批量提取掃描版PDF需要轉(zhuǎn)換為可編輯文本不同格式的圖片需要統(tǒng)一轉(zhuǎn)換格式自動(dòng)化方案的優(yōu)勢(shì)批量處理一次性處理數(shù)百個(gè)文件一致性保證所有文件采用相同處理參數(shù)時(shí)間節(jié)省從小時(shí)級(jí)縮短到分鐘級(jí)可重復(fù)使用處理邏輯可封裝為腳本重復(fù)調(diào)用以技術(shù)文檔為例一個(gè)中等規(guī)模的項(xiàng)目可能包含50-100張示意圖、架構(gòu)圖等圖片資源。手動(dòng)處理每張圖片平均需要2-3分鐘而自動(dòng)化腳本可以在30秒內(nèi)完成全部處理。2. 環(huán)境準(zhǔn)備與工具選擇2.1 基礎(chǔ)環(huán)境要求# 檢查Python版本 python --version # 推薦 Python 3.82.2 核心依賴庫(kù)安裝pip install Pillow PyPDF2 pdf2image python-docx pip install opencv-python pytesseract2.3 各庫(kù)的功能說(shuō)明庫(kù)名稱主要功能適用場(chǎng)景Pillow圖像處理核心庫(kù)圖片縮放、格式轉(zhuǎn)換、濾鏡應(yīng)用PyPDF2PDF文本處理PDF拆分、合并、文本提取pdf2imagePDF轉(zhuǎn)圖片將PDF頁(yè)面轉(zhuǎn)換為圖像格式python-docxWord文檔操作處理提取的文本內(nèi)容OpenCV高級(jí)圖像處理圖像識(shí)別、邊緣檢測(cè)pytesseractOCR文字識(shí)別從圖片中提取文字2.4 環(huán)境驗(yàn)證腳本# environment_check.py import importlib required_libraries [PIL, PyPDF2, pdf2image, cv2, pytesseract] def check_environment(): missing_libs [] for lib in required_libraries: try: importlib.import_module(lib) print(f? {lib} 安裝成功) except ImportError: missing_libs.append(lib) print(f? {lib} 未安裝) if missing_libs: print(f\n需要安裝的庫(kù): {, .join(missing_libs)}) return False return True if __name__ __main__: check_environment()3. 圖片批量處理實(shí)戰(zhàn)3.1 基礎(chǔ)圖片處理類設(shè)計(jì)# image_processor.py from PIL import Image import os from pathlib import Path class ImageProcessor: def __init__(self, source_dir, output_dir): self.source_dir Path(source_dir) self.output_dir Path(output_dir) self.output_dir.mkdir(exist_okTrue) def get_supported_formats(self): 獲取支持的圖片格式 return [.jpg, .jpeg, .png, .bmp, .tiff, .webp] def find_images(self): 查找目錄中的所有圖片文件 images [] for format_ext in self.get_supported_formats(): images.extend(self.source_dir.glob(f*{format_ext})) images.extend(self.source_dir.glob(f*{format_ext.upper()})) return images3.2 圖片尺寸批量調(diào)整# 續(xù) image_processor.py def resize_images(self, target_size(800, 600), keep_aspectTrue): 批量調(diào)整圖片尺寸 images self.find_images() processed_count 0 for img_path in images: try: with Image.open(img_path) as img: if keep_aspect: # 保持寬高比調(diào)整尺寸 img.thumbnail(target_size, Image.Resampling.LANCZOS) else: # 強(qiáng)制調(diào)整到指定尺寸 img img.resize(target_size, Image.Resampling.LANCZOS) # 保存處理后的圖片 output_path self.output_dir / fresized_{img_path.name} img.save(output_path, optimizeTrue) processed_count 1 print(f已處理: {img_path.name} - {output_path.name}) except Exception as e: print(f處理失敗 {img_path.name}: {str(e)}) return processed_count3.3 圖片格式批量轉(zhuǎn)換# 續(xù) image_processor.py def convert_format(self, target_formatJPEG, quality85): 批量轉(zhuǎn)換圖片格式 images self.find_images() converted_count 0 for img_path in images: try: with Image.open(img_path) as img: # 轉(zhuǎn)換為RGB模式針對(duì)JPEG格式 if img.mode ! RGB and target_format JPEG: img img.convert(RGB) output_filename f{img_path.stem}.{target_format.lower()} output_path self.output_dir / output_filename save_kwargs {quality: quality} if target_format JPEG else {} img.save(output_path, formattarget_format, **save_kwargs) converted_count 1 print(f已轉(zhuǎn)換: {img_path.name} - {output_filename}) except Exception as e: print(f轉(zhuǎn)換失敗 {img_path.name}: {str(e)}) return converted_count3.4 圖片壓縮優(yōu)化# 續(xù) image_processor.py def compress_images(self, quality70, max_sizeNone): 批量壓縮圖片 images self.find_images() compressed_count 0 total_saved 0 for img_path in images: try: original_size img_path.stat().st_size with Image.open(img_path) as img: # 如果有最大尺寸限制先調(diào)整尺寸 if max_size: img.thumbnail(max_size, Image.Resampling.LANCZOS) output_path self.output_dir / fcompressed_{img_path.name} # 根據(jù)格式選擇保存參數(shù) if img_path.suffix.lower() in [.jpg, .jpeg]: img.save(output_path, optimizeTrue, qualityquality) else: img.save(output_path, optimizeTrue) compressed_size output_path.stat().st_size saved original_size - compressed_size total_saved saved compressed_count 1 print(f壓縮: {img_path.name} f({original_size//1024}KB - {compressed_size//1024}KB) f節(jié)省: {saved//1024}KB) except Exception as e: print(f壓縮失敗 {img_path.name}: {str(e)}) print(f\n總計(jì)壓縮 {compressed_count} 張圖片節(jié)省空間: {total_saved//1024}KB) return compressed_count, total_saved4. PDF處理全流程實(shí)現(xiàn)4.1 PDF基礎(chǔ)操作類# pdf_processor.py import PyPDF2 from pdf2image import convert_from_path import pytesseract from PIL import Image import os class PDFProcessor: def __init__(self, poppler_pathNone): self.poppler_path poppler_path def split_pdf(self, input_pdf, output_dir, pages_per_split10): 拆分PDF文件 with open(input_pdf, rb) as file: pdf_reader PyPDF2.PdfReader(file) total_pages len(pdf_reader.pages) for start_page in range(0, total_pages, pages_per_split): end_page min(start_page pages_per_split, total_pages) pdf_writer PyPDF2.PdfWriter() for page_num in range(start_page, end_page): pdf_writer.add_page(pdf_reader.pages[page_num]) output_filename f{os.path.splitext(input_pdf)[0]}_part{start_page//pages_per_split 1}.pdf output_path os.path.join(output_dir, output_filename) with open(output_path, wb) as output_file: pdf_writer.write(output_file) print(f生成拆分文件: {output_filename} (頁(yè)碼 {start_page1}-{end_page}))4.2 PDF轉(zhuǎn)圖片處理# 續(xù) pdf_processor.py def pdf_to_images(self, input_pdf, output_dir, dpi200): 將PDF轉(zhuǎn)換為圖片 os.makedirs(output_dir, exist_okTrue) try: images convert_from_path(input_pdf, dpidpi, poppler_pathself.poppler_path) for i, image in enumerate(images): output_path os.path.join(output_dir, fpage_{i1:03d}.jpg) image.save(output_path, JPEG, quality85) print(f轉(zhuǎn)換頁(yè)面 {i1} - {output_path}) return len(images) except Exception as e: print(fPDF轉(zhuǎn)換圖片失敗: {str(e)}) return 04.3 PDF文字提取與OCR# 續(xù) pdf_processor.py def extract_text(self, input_pdf, use_ocrFalse): 提取PDF中的文字內(nèi)容 text_content try: # 首先嘗試直接提取文本 with open(input_pdf, rb) as file: pdf_reader PyPDF2.PdfReader(file) for page_num in range(len(pdf_reader.pages)): page_text pdf_reader.pages[page_num].extract_text() if page_text.strip(): text_content f--- 第 {page_num1} 頁(yè) ---\n{page_text}\n\n # 如果直接提取文本較少且啟用OCR嘗試OCR識(shí)別 if use_ocr and len(text_content.strip()) 100: print(文本提取較少啟用OCR識(shí)別...) ocr_text self._ocr_pdf(input_pdf) text_content \n--- OCR識(shí)別結(jié)果 ---\n ocr_text except Exception as e: print(f文本提取失敗: {str(e)}) return text_content def _ocr_pdf(self, input_pdf): 使用OCR識(shí)別PDF中的文字 ocr_text temp_dir temp_ocr os.makedirs(temp_dir, exist_okTrue) try: # 先將PDF轉(zhuǎn)換為圖片 image_count self.pdf_to_images(input_pdf, temp_dir, dpi300) # 對(duì)每張圖片進(jìn)行OCR識(shí)別 for i in range(image_count): image_path os.path.join(temp_dir, fpage_{i1:03d}.jpg) if os.path.exists(image_path): text pytesseract.image_to_string(Image.open(image_path), langchi_simeng) ocr_text f第 {i1} 頁(yè):\n{text}\n\n # 清理臨時(shí)文件 for file in os.listdir(temp_dir): os.remove(os.path.join(temp_dir, file)) os.rmdir(temp_dir) except Exception as e: print(fOCR識(shí)別失敗: {str(e)}) return ocr_text5. 完整工作流整合5.1 自動(dòng)化處理管道# workflow_manager.py from image_processor import ImageProcessor from pdf_processor import PDFProcessor import os from datetime import datetime class DocumentWorkflow: def __init__(self, base_workspaceworkspace): self.workspace base_workspace self.setup_workspace() def setup_workspace(self): 創(chuàng)建工作區(qū)目錄結(jié)構(gòu) directories [ source_images, source_pdfs, processed_images, processed_pdfs, output_docs, temp ] for dir_name in directories: os.makedirs(os.path.join(self.workspace, dir_name), exist_okTrue) def run_image_processing_pipeline(self, resize_to(1200, 800), compress_quality80): 運(yùn)行圖片處理管道 print(開始圖片批量處理...) processor ImageProcessor( os.path.join(self.workspace, source_images), os.path.join(self.workspace, processed_images) ) # 執(zhí)行處理流程 resize_count processor.resize_images(resize_to) compress_count, saved_space processor.compress_images(compress_quality) print(f圖片處理完成: 調(diào)整尺寸 {resize_count} 張, 壓縮 {compress_count} 張) return resize_count, compress_count, saved_space def run_pdf_processing_pipeline(self, enable_ocrTrue): 運(yùn)行PDF處理管道 print(開始PDF批量處理...) pdf_source_dir os.path.join(self.workspace, source_pdfs) pdf_files [f for f in os.listdir(pdf_source_dir) if f.lower().endswith(.pdf)] processor PDFProcessor() total_text_length 0 for pdf_file in pdf_files: pdf_path os.path.join(pdf_source_dir, pdf_file) print(f\n處理PDF: {pdf_file}) # 提取文本 text_content processor.extract_text(pdf_path, use_ocrenable_ocr) total_text_length len(text_content) # 保存提取的文本 text_output_path os.path.join( self.workspace, output_docs, f{os.path.splitext(pdf_file)[0]}_extracted.txt ) with open(text_output_path, w, encodingutf-8) as f: f.write(text_content) print(f文本已保存: {text_output_path}) print(fPDF處理完成: 共處理 {len(pdf_files)} 個(gè)文件, 提取文本 {total_text_length} 字符) return len(pdf_files), total_text_length5.2 批處理腳本示例# batch_processor.py #!/usr/bin/env python3 圖片和PDF批量處理腳本 使用方法: python batch_processor.py --images --pdfs --workspace ./my_docs import argparse import sys from workflow_manager import DocumentWorkflow def main(): parser argparse.ArgumentParser(description文檔批量處理工具) parser.add_argument(--images, actionstore_true, help處理圖片) parser.add_argument(--pdfs, actionstore_true, help處理PDF) parser.add_argument(--workspace, defaultworkspace, help工作目錄) parser.add_argument(--resize, nargs2, typeint, default[1200, 800], help圖片目標(biāo)尺寸 寬 高) parser.add_argument(--quality, typeint, default80, help壓縮質(zhì)量) args parser.parse_args() if not args.images and not args.pdfs: print(請(qǐng)指定處理類型: --images 或 --pdfs) sys.exit(1) # 初始化工作流 workflow DocumentWorkflow(args.workspace) results {} # 執(zhí)行圖片處理 if args.images: print( * 50) print(開始圖片批量處理流程) print( * 50) resize_count, compress_count, saved_space workflow.run_image_processing_pipeline( tuple(args.resize), args.quality ) results[images] { resized: resize_count, compressed: compress_count, saved_space_kb: saved_space // 1024 } # 執(zhí)行PDF處理 if args.pdfs: print( * 50) print(開始PDF批量處理流程) print( * 50) pdf_count, text_length workflow.run_pdf_processing_pipeline() results[pdfs] { processed: pdf_count, text_extracted: text_length } # 輸出處理報(bào)告 print(\n * 50) print(處理完成報(bào)告) print( * 50) for task_type, stats in results.items(): print(f\n{task_type.upper()} 處理結(jié)果:) for stat_name, value in stats.items(): print(f {stat_name}: {value}) if __name__ __main__: main()6. 實(shí)戰(zhàn)案例技術(shù)文檔整理6.1 場(chǎng)景描述假設(shè)我們有一個(gè)技術(shù)項(xiàng)目包含以下文件50張不同尺寸的架構(gòu)圖、流程圖截圖3份掃描版的技術(shù)規(guī)范PDF文檔需要統(tǒng)一處理為標(biāo)準(zhǔn)化格式6.2 具體操作步驟# 1. 準(zhǔn)備目錄結(jié)構(gòu) mkdir -p my_project/{source_images,source_pdfs} # 2. 將文件放入對(duì)應(yīng)目錄 # source_images/ 放入所有圖片文件 # source_pdfs/ 放入所有PDF文件 # 3. 運(yùn)行處理腳本 python batch_processor.py --images --pdfs --workspace my_project --resize 1000 800 --quality 856.3 處理結(jié)果驗(yàn)證處理完成后檢查生成的文件my_project/processed_images/調(diào)整尺寸和壓縮后的圖片my_project/output_docs/從PDF提取的文本內(nèi)容7. 常見問(wèn)題與解決方案7.1 圖片處理常見問(wèn)題問(wèn)題現(xiàn)象可能原因解決方案圖片處理后顏色失真色彩模式不匹配轉(zhuǎn)換時(shí)確保使用RGB模式透明背景變成黑色格式轉(zhuǎn)換問(wèn)題PNG轉(zhuǎn)JPEG時(shí)處理透明度文件大小反而變大壓縮參數(shù)設(shè)置不當(dāng)調(diào)整quality參數(shù)通常70-85為宜處理速度過(guò)慢圖片尺寸過(guò)大先縮小尺寸再進(jìn)行處理7.2 PDF處理常見問(wèn)題問(wèn)題現(xiàn)象可能原因解決方案中文OCR識(shí)別率低語(yǔ)言包未安裝安裝中文語(yǔ)言包c(diǎn)hi_simPDF轉(zhuǎn)換圖片失敗poppler路徑錯(cuò)誤指定正確的poppler路徑文本提取為空掃描版PDF啟用OCR功能內(nèi)存不足錯(cuò)誤PDF頁(yè)數(shù)過(guò)多分批次處理7.3 環(huán)境配置問(wèn)題排查# troubleshooting.py def diagnose_common_issues(): 診斷常見環(huán)境問(wèn)題 issues [] # 檢查Poppler路徑 try: from pdf2image import convert_from_path test_pdf test.pdf if os.path.exists(test_pdf): convert_from_path(test_pdf, first_page1, last_page1) except Exception as e: issues.append(fPDF轉(zhuǎn)換問(wèn)題: {e}) # 檢查Tesseract OCR try: import pytesseract pytesseract.get_tesseract_version() except Exception as e: issues.append(fOCR引擎問(wèn)題: {e}) return issues8. 性能優(yōu)化與最佳實(shí)踐8.1 內(nèi)存優(yōu)化策略# optimized_processor.py class OptimizedImageProcessor(ImageProcessor): def __init__(self, source_dir, output_dir, max_memory_mb500): super().__init__(source_dir, output_dir) self.max_memory_mb max_memory_mb def process_large_batch(self, batch_size10): 分批處理大文件集避免內(nèi)存溢出 all_images self.find_images() for i in range(0, len(all_images), batch_size): batch all_images[i:i batch_size] print(f處理批次 {i//batch_size 1}/{(len(all_images)-1)//batch_size 1}) for img_path in batch: self._process_single_image(img_path) # 手動(dòng)觸發(fā)垃圾回收 import gc gc.collect()8.2 多線程處理加速# parallel_processor.py import concurrent.futures from functools import partial class ParallelProcessor: def __init__(self, max_workers4): self.max_workers max_workers def parallel_image_processing(self, image_paths, process_function): 并行處理圖片 with concurrent.futures.ThreadPoolExecutor(max_workersself.max_workers) as executor: # 使用partial固定其他參數(shù) process_func partial(process_function) results list(executor.map(process_func, image_paths)) return results8.3 配置文件管理# config_manager.py import json from pathlib import Path class ConfigManager: def __init__(self, config_fileprocessing_config.json): self.config_file Path(config_file) self.default_config { image_processing: { target_size: [1200, 800], compression_quality: 80, keep_aspect_ratio: True }, pdf_processing: { ocr_enabled: True, dpi: 200, language: chi_simeng } } def load_config(self): 加載配置文件 if self.config_file.exists(): with open(self.config_file, r, encodingutf-8) as f: return json.load(f) else: self.save_config(self.default_config) return self.default_config def save_config(self, config): 保存配置文件 with open(self.config_file, w, encodingutf-8) as f: json.dump(config, f, indent2, ensure_asciiFalse)9. 擴(kuò)展功能與自定義開發(fā)9.1 添加水印功能# watermark_processor.py from PIL import Image, ImageDraw, ImageFont class WatermarkProcessor: def add_watermark_batch(self, image_dir, watermark_text, positionbottom-right): 批量添加水印 processor ImageProcessor(image_dir, image_dir _watermarked) images processor.find_images() for img_path in images: self._add_single_watermark(img_path, watermark_text, position) def _add_single_watermark(self, image_path, text, position): 為單張圖片添加水印 with Image.open(image_path) as img: # 創(chuàng)建水印層 watermark Image.new(RGBA, img.size, (0, 0, 0, 0)) draw ImageDraw.Draw(watermark) # 計(jì)算水印位置 bbox draw.textbbox((0, 0), text) text_width bbox[2] - bbox[0] text_height bbox[3] - bbox[1] if position bottom-right: x img.width - text_width - 20 y img.height - text_height - 20 elif position center: x (img.width - text_width) // 2 y (img.height - text_height) // 2 # 繪制水印文字 draw.text((x, y), text, fill(255, 255, 255, 128)) # 合并圖片和水印 watermarked Image.alpha_composite(img.convert(RGBA), watermark) watermarked watermarked.convert(RGB) # 保存結(jié)果 output_path self.output_dir / fwatermarked_{image_path.name} watermarked.save(output_path, quality85)9.2 批量重命名與元數(shù)據(jù)處理# metadata_processor.py from PIL.ExifTags import TAGS import os from datetime import datetime class MetadataProcessor: def rename_by_date(self, image_dir, pattern{date}_{counter}): 按拍攝日期重命名圖片 images self.find_images(image_dir) for counter, img_path in enumerate(images, 1): try: with Image.open(img_path) as img: exif_data img._getexif() date_taken None if exif_data: for tag_id, value in exif_data.items(): tag TAGS.get(tag_id, tag_id) if tag DateTime: date_taken value.replace(:, ).replace( , _) break # 如果沒(méi)有EXIF數(shù)據(jù)使用文件修改時(shí)間 if not date_taken: mtime os.path.getmtime(img_path) date_taken datetime.fromtimestamp(mtime).strftime(%Y%m%d_%H%M%S) new_name pattern.format(datedate_taken, countercounter) new_path img_path.parent / f{new_name}{img_path.suffix} os.rename(img_path, new_path) print(f重命名: {img_path.name} - {new_path.name}) except Exception as e: print(f重命名失敗 {img_path.name}: {str(e)})這套圖片和PDF批量處理方案在實(shí)際項(xiàng)目中經(jīng)過(guò)驗(yàn)證能夠?qū)⒃拘枰獢?shù)小時(shí)的手工操作壓縮到幾分鐘內(nèi)完成。關(guān)鍵在于根據(jù)具體需求調(diào)整參數(shù)并建立標(biāo)準(zhǔn)化的處理流程。建議在實(shí)際使用前先用少量測(cè)試文件驗(yàn)證處理效果確認(rèn)符合預(yù)期后再進(jìn)行批量處理。對(duì)于重要的原始文件務(wù)必先做好備份避免處理過(guò)程中出現(xiàn)意外情況導(dǎo)致數(shù)據(jù)丟失。