result_output_v4_2023.09.13.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 2,
   "id": "91c5912b",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\u001b[34m\u001b[1mdetect: \u001b[0mweights=['C:\\\\Users\\\\sj990\\\\MachineLearning\\\\CAPSTONE_DESIGN_AI_Module\\\\yolo\\\\best_noodle_drink.pt'], source=C:\\Users\\sj990\\MachineLearning\\CAPSTONE_DESIGN_AI_Module\\yolo\\first_image_folder - \\KakaoTalk_20230918_100236376.jpg, data=yolov5\\data\\coco128.yaml, imgsz=[640, 640], conf_thres=0.1, iou_thres=0.45, max_det=1000, device=, view_img=False, save_txt=False, save_conf=False, save_crop=True, nosave=False, classes=None, agnostic_nms=False, augment=False, visualize=False, update=False, project=../yolo, name=exp, exist_ok=False, line_thickness=3, hide_labels=False, hide_conf=False, half=False, dnn=False, vid_stride=1\n",
      "YOLOv5  v7.0-198-g34c2187 Python-3.9.7 torch-2.0.0+cpu CPU\n",
      "\n",
      "Fusing layers... \n",
      "Model summary: 157 layers, 7015519 parameters, 0 gradients, 15.8 GFLOPs\n",
      "image 1/1 C:\\Users\\sj990\\MachineLearning\\CAPSTONE_DESIGN_AI_Module\\yolo\\first_image_folder - \\KakaoTalk_20230918_100236376.jpg: 640x480 2 noodles, 128.6ms\n",
      "Speed: 1.0ms pre-process, 128.6ms inference, 1.0ms NMS per image at shape (1, 3, 640, 640)\n",
      "Results saved to \u001b[1m..\\yolo\\exp\u001b[0m\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "WARNING:tensorflow:No training configuration found in the save file, so the model was *not* compiled. Compile it manually.\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "WARNING:tensorflow:No training configuration found in the save file, so the model was *not* compiled. Compile it manually.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "1/1 [==============================] - 0s 378ms/step\n",
      "[{'상품': '65053_엠디에스코리아베이컨까르보나라파스타', '정확도': 0.32}]\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "' 실사용시 주석제거바람\\nfor file in os.scandir(application_image):\\n    os.remove(file.path)\\n'"
      ]
     },
     "execution_count": 2,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "import tensorflow as tf\n",
    "import tensorflow_hub as hub\n",
    "import numpy as np\n",
    "import glob     \n",
    "import json\n",
    "import os\n",
    "import shutil\n",
    "\n",
    "#########################################################################\n",
    "#########################################################################\n",
    "# 수정해야되는 초기경로설정들 \n",
    "# 1. 어플에서 올라온 이미지폴더 경로\n",
    "application_image = r'C:\\Users\\sj990\\MachineLearning\\CAPSTONE_DESIGN_AI_Module\\yolo\\first_image_folder - 복사본'\n",
    "\n",
    "# 2. yolov5모델 경로 \n",
    "yolo_model = r'C:\\Users\\sj990\\MachineLearning\\CAPSTONE_DESIGN_AI_Module\\yolo\\best_noodle_drink.pt'\n",
    "\n",
    "# 2. 면류 모델, 라벨경로\n",
    "noodle_model_path = r'C:\\Users\\sj990\\Desktop\\c_d\\model_label\\models\\model_noodle.h5'\n",
    "noodle_label_path = r'C:\\Users\\sj990\\Desktop\\c_d\\model_label\\labels\\model_noodle_label.txt'\n",
    "\n",
    "# 3. 음료 모델, 라벨경로\n",
    "drink_model_path = r'C:\\Users\\sj990\\Desktop\\c_d\\model_label\\models\\model_drink.h5'\n",
    "drink_label_path = r'C:\\Users\\sj990\\Desktop\\c_d\\model_label\\labels\\model_drink_label.txt'\n",
    "\n",
    "#4 과자 모델, 라벨경로\n",
    "snack_model_path = ''\n",
    "snack_label_path = ''\n",
    "\n",
    "\n",
    "#5 생활용품 모델, 라벨경로\n",
    "house_model_path = ''\n",
    "house_label_path = ''\n",
    "\n",
    "\n",
    "# 미분류 -> 지울거\n",
    "unclassified_model_path = r'C:\\Users\\sj990\\Desktop\\종설\\모델,라벨\\모델\\model_unclass.h5'\n",
    "unclassified_label_path = r'C:\\Users\\sj990\\Desktop\\종설\\모델,라벨\\라벨\\model_unclass_label.txt'\n",
    "\n",
    "#########################################################################\n",
    "########################################################################\n",
    "\n",
    "\n",
    "image_directory = application_image\n",
    "# jpg, png, jpeg 형식 (리스트형식으로 저장된다, 경로에는 파일한장만 존재하게)\n",
    "image_files = glob.glob(image_directory + '/*.png') + glob.glob(image_directory + '/*.jpg') + glob.glob(image_directory + '/*.jpeg')\n",
    "# 리스트중 첫번째것만 사용\n",
    "first_image = image_files[0]\n",
    "\n",
    "# 객체검출 실행                                                conf조절하면서 객체검출\n",
    "!python yolov5/detect.py --weights \"{yolo_model}\" --img 640 --conf 0.1 --source \"{first_image}\" --project ../yolo --save-crop\n",
    "\n",
    "# 검출경로\n",
    "crops_path = os.path.join('exp', 'crops')\n",
    "\n",
    "# 객체 폴더 있는지 탐지 (매번 추가) - 작업중, 생활용품은 3가지에서 검출안될때 (else)\n",
    "noodle_detect = os.path.exists(os.path.join(crops_path, 'noodle'))\n",
    "drink_detect = os.path.exists(os.path.join(crops_path, 'drink'))\n",
    "snack_detect = os.path.exists(os.path.join(crops_path, 'snack'))\n",
    "\n",
    "\n",
    "###############################################################\n",
    "#내가 하면서 추가\n",
    "if noodle_detect:\n",
    "    yolo_img = os.path.join(crops_path, 'noodle')\n",
    "    classification_label = 'noodle'\n",
    "    \n",
    "elif drink_detect:\n",
    "    yolo_img = os.path.join(crops_path, 'drink')\n",
    "    classification_label = 'drink'\n",
    "    \n",
    "elif snack_detect:\n",
    "    yolo_img = os.path.join(crops_path, 'snack')\n",
    "    classification_label = 'snack'\n",
    "    \n",
    "else:\n",
    "    yolo_img = application_image\n",
    "    classification_label = 'house'\n",
    "###############################################################\n",
    "    \n",
    "    \n",
    "yolo_image_directory = yolo_img\n",
    "# jpg, png, jpeg 형식 (리스트형식으로 저장된다, 경로에는 파일한장만 존재하게)\n",
    "yolo_image_files = glob.glob(yolo_image_directory + '/*.png') + glob.glob(yolo_image_directory + '/*.jpg') + glob.glob(yolo_image_directory + '/*.jpeg')\n",
    "# 리스트중 첫번째것만 사용\n",
    "yolo_image_path = yolo_image_files[0]\n",
    "\n",
    "\n",
    "#####################모델정의#####################\n",
    "class Model:\n",
    "    # 초기화\n",
    "    def __init__(self, image_path, model_path, class_labels_path):\n",
    "        self.image_path = image_path\n",
    "        self.model_path = model_path\n",
    "        self.class_labels_path = class_labels_path\n",
    "        \n",
    "    #예측 (전처리된이미지와 클래스라벨을 삽입)\n",
    "    def preds_model(self, img, class_labels):\n",
    "        model = tf.keras.models.load_model(self.model_path,  custom_objects={'KerasLayer': hub.KerasLayer})\n",
    "        preds = model.predict(img) \n",
    "        \n",
    "        # 결과값 상위 1개\n",
    "        top_preds_idx = preds[0].argsort()[ : : -1][ : 1]\n",
    "        top_preds_labels = [class_labels[idx] for idx in top_preds_idx]\n",
    "        top_preds_probs = preds[0][top_preds_idx]\n",
    "        return top_preds_labels, top_preds_probs\n",
    "    \n",
    "    #이미지전처리_1 (기본)\n",
    "    def img_pre_default(self):  \n",
    "        img = tf.keras.preprocessing.image.load_img(self.image_path, target_size=(256, 256))\n",
    "        img = tf.keras.preprocessing.image.img_to_array(img)\n",
    "        img = img.astype('float32') / 255.0\n",
    "        img = np.expand_dims(img, axis=0)\n",
    "        return img\n",
    "\n",
    "    #라벨불러오기\n",
    "    def class_label(self):\n",
    "        class_labels = []\n",
    "        with open(self.class_labels_path, 'r') as file:\n",
    "            for line in file:\n",
    "                class_labels.append(line.strip())\n",
    "        return class_labels\n",
    "    \n",
    "    #결과출력\n",
    "    def print_result(self, top_preds_labels, top_preds_probs):\n",
    "        results = []\n",
    "        #json형식 results\n",
    "        for label, prob in zip(top_preds_labels, top_preds_probs):\n",
    "            result = {\n",
    "                '상품': label,\n",
    "                '정확도': round(float(prob), 2)\n",
    "            }\n",
    "            results.append(result)\n",
    "        print(results)\n",
    "        return results\n",
    "#####################모델정의#####################\n",
    "\n",
    "\n",
    "image_path = yolo_image_path\n",
    "\n",
    "###########특징분류함수###########\n",
    "def save_result(classification_label, image_path, model_path, label_path):\n",
    "    model = Model(image_path, model_path, label_path)\n",
    "    \n",
    "    #예측\n",
    "    result_labels, result_probs = model.preds_model(model.img_pre_default(), model.class_label())\n",
    "    \n",
    "    #결과출력\n",
    "    result = model.print_result(result_labels, result_probs)\n",
    "    \n",
    "    # json파일 저장\n",
    "    with open('results.json', 'w', encoding='utf-8') as file:\n",
    "        json.dump(result, file, ensure_ascii=False)\n",
    "\n",
    "############특징분류############\n",
    "if classification_label == 'noodle':\n",
    "    save_result(\"noodle\", image_path, noodle_model_path, noodle_label_path)\n",
    "    \n",
    "elif classification_label == \"drink\":\n",
    "    save_result(\"drink\", image_path, drink_model_path, drink_label_path)\n",
    "    \n",
    "elif classification_label == \"snack\":\n",
    "    save_result(\"snack\", image_path, drink_model_path, drink_label_path) \n",
    "    \n",
    "elif classification_label == 'house':\n",
    "    print('아직생활용품은작업중')\n",
    "    #save_result(\"house\", image_path, drink_model_path, drink_label_path) \n",
    "    \n",
    "elif classification_label == \"Unclassified\":\n",
    "    save_result(\"Unclassified\", image_path, unclassified_model_path, unclassified_label_path)\n",
    "###############################      \n",
    "\n",
    "    \n",
    "# 작업끝난후 yolo 이미지 폴더 전체 삭제\n",
    "#shutil.rmtree(\"exp\")\n",
    "\n",
    "# 작업끝난후 어플리케이션에서 올라온 이미지만 삭제\n",
    "\"\"\" 실사용시 주석제거바람\n",
    "for file in os.scandir(application_image):\n",
    "    os.remove(file.path)\n",
    "\"\"\""
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "765fe827",
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "python39",
   "language": "python",
   "name": "python39"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.9.16"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}