Revert "paddleocr"

This reverts commit f6e9138688ef632268208c2a6b0fb4ae7c5924fc.

Dockerfile

@@ -1,11 +1,10 @@
 FROM python:3.12.1
 
 ENV HF_HOME /tmp
-ENV HOME /tmp
 
 WORKDIR /code
 
-RUN apt-get update && apt-get install -y libgl1-mesa-glx libboost-thread1.74.0 libxext6 fonts-freefont-ttf && rm -rf /var/lib/apt/lists/*
+RUN apt-get update && apt-get install -y libgl1-mesa-glx libboost-thread1.74.0 && rm -rf /var/lib/apt/lists/*
 
 
 

main.py

@@ -6,11 +6,11 @@ from datetime import datetime
 from typing import List
 
 import cv2
+import easyocr
 import numpy as np
 from fastapi import FastAPI, HTTPException, Request, UploadFile
 from fastapi.middleware.cors import CORSMiddleware
 from fastapi.responses import JSONResponse
-from paddleocr import PaddleOCR
 from PIL import Image
 
 import dds
@@ -62,18 +62,18 @@ def load_ocr_model():
     グローバル変数readerに格納する。
     """
     global reader
-    # model_dir = "/tmp/easyocr_models"
-    # network_dir = "/tmp/easyocr_models/network"
+    model_dir = "/tmp/easyocr_models"
+    network_dir = "/tmp/easyocr_models/network"
 
     # 使用する言語と、モデルの保存先ディレクトリを指定してReaderを初期化
-    # reader = easyocr.Reader(
-    #     ["en"],
-    #     model_storage_directory=model_dir,
-    #     user_network_directory=network_dir,
-    # )
-
-    reader = PaddleOCR(use_angle_cls=True, lang="en", show_log=False)
-    print("PaddleOCR model loaded successfully.")
+    reader = easyocr.Reader(
+        ["en"],
+        model_storage_directory=model_dir,
+        user_network_directory=network_dir,
+    )
+    print(
+        f"EasyOCR model loaded successfully. Models are stored in {model_dir}"
+    )
 
 
 # @app.on_event("startup")
@@ -213,11 +213,14 @@ async def analyze_image(image_paths: list[UploadFile]):
                 # img = Image.fromarray(
                 #     cv2.cvtColor(candidate["img"], cv2.COLOR_BGR2RGB)
                 # )
+                custom_config = (
+                    r"--psm 10 -c tessedit_char_whitelist=0123456789AKQJ"
+                )
 
-                result = reader.ocr(candidate["img"], cls=True)
-                # result = reader.readtext(
-                #     candidate["img"], text_threshold=0.4
-                # )
+                # result = reader.readtext(candidate["img"])
+                result = reader.readtext(
+                    candidate["img"], text_threshold=0.4
+                )
                 print(result)
                 if len(result) > 0:
                     ocr_results.append(result)
@@ -239,7 +242,7 @@ async def analyze_image(image_paths: list[UploadFile]):
         # reds = []
         for i, result in enumerate(ocr_results):
             # text = result[0]["generated_text"].upper().strip()
-            text = result[0][1][0]
+            _, text, _ = result[0]
             print(text, is_text_valid(text))
 
             text = is_text_valid(text)