Merge pull request #36 from kookmin-sw/feat/flask_data_extract

Feat/flask data extract

.gitignore

@@ -9,3 +9,4 @@ personal_color_dataset/valid/img*.jpg
 gan/pretrained_models
 gan/output
 predict_image/*
+

Color_extract/color.py

@@ -7,6 +7,7 @@
 sys.path.append(os.path.dirname(os.path.abspath(os.path.dirname(__file__))))
 from Skin_detect.skin_detect_v2 import *
 from image_processing.gamma_correction import *
+from PC_model.utils import draw_3d_rgb
 
 def extract_points(mask, img):
     points = np.argwhere(mask)
@@ -256,7 +257,9 @@ def extract_high_rank(rgb_codes, color_area, percent):
 #     # file 이름 넣기
 #     data['filename'][i] = name
 
-def total_data_extract(path):
+def total_data_extract(path, save_image):
+    folder_path = os.path.dirname(path)
+
     data = {'Red' : 0, 'Green' : 0, 'Blue' : 0, 
         'Hue' : 0, 'Saturation' : 0, 'Value' : 0, 
         'Y' : 0, 'Cr' : 0, 'Cb' : 0, 
@@ -278,6 +281,10 @@ def total_data_extract(path):
     # 이진 마스크로 변환
     binary_mask = (face_nose_mask >= 0.5).astype(int)
 
+    if save_image:
+        if not os.path.exists(os.path.join(folder_path, "binary_mask.jpg")):
+            cv2.imwrite(os.path.join(folder_path, "binary_mask.jpg"), binary_mask * 255)
+
     # image load
     image = cv2.imread(path)
     image = gamma_correction(image, 0.8)
@@ -286,6 +293,12 @@ def total_data_extract(path):
     ycrcb_image = cv2.cvtColor(image, cv2.COLOR_BGR2YCrCb)
     lab_image = cv2.cvtColor(image, cv2.COLOR_BGR2LAB)
 
+    if save_image:
+        if not os.path.exists(os.path.join(folder_path, "face_nose_img.jpg")):
+            face_nose_img = np.zeros_like(image)
+            face_nose_img[binary_mask == 1] = image[binary_mask == 1]
+            cv2.imwrite(os.path.join(folder_path, "face_nose_img.jpg"), face_nose_img)
+
     #RGB
     rgb = extract_points(binary_mask, rgb_image)
     rgb_average = rgb.mean(axis=0).round()
@@ -348,4 +361,8 @@ def total_data_extract(path):
     data['New Green'] = new_rgb_average[1]
     data['New Blue'] = new_rgb_average[2]
 
-    return data
+    if save_image:
+        draw_3d_rgb(new_rgb_codes, new_rgb_codes/255.0, folder_path)
+
+    return data
+

PC_model/utils.py

@@ -1,7 +1,10 @@
 from sklearn.metrics import confusion_matrix, accuracy_score, precision_score, recall_score
 from sklearn.metrics import confusion_matrix, f1_score
 import seaborn as sns
+import matplotlib
+matplotlib.use("agg")
 import matplotlib.pyplot as plt
+import os
 
 
 def get_evaluation(y_test, y_pred):
@@ -28,3 +31,18 @@ def feature_plot(data, label_name):
 
 def heatmap_plot(data, number = True):
     sns.heatmap(data, annot=number, fmt=".2f")
+
+# 3d plot을 이용하여 데이터 분포 확인
+def draw_3d_rgb(rgb_codes, colors, path = None):
+    fig = plt.figure()
+    ax = fig.add_subplot(111, projection='3d')
+    ax.scatter(rgb_codes[:, 0], rgb_codes[:, 1], rgb_codes[:, 2], c=colors, marker='o')
+    ax.set_xlabel('Red')
+    ax.set_ylabel('Green')
+    ax.set_zlabel('Blue')
+    if path is None:
+        plt.show()
+    else:
+        if not os.path.exists(path + "/rgb_3d_plot.jpg"):
+            plt.savefig(path + "/rgb_3d_plot.jpg", format='jpeg')
+

backend/build.gradle

@@ -25,6 +25,7 @@ dependencies {
     implementation 'org.springframework.boot:spring-boot-starter-thymeleaf'
     implementation 'org.springframework.boot:spring-boot-starter-web'
     implementation 'org.springframework.boot:spring-boot-starter-data-jpa'
+    implementation 'com.google.code.gson:gson:2.8.6'
     compileOnly 'org.projectlombok:lombok'
     compileOnly'org.springframework.boot:spring-boot-devtools'
     developmentOnly 'org.springframework.boot:spring-boot-devtools'

backend/src/main/java/org/capstone2024/onlyu/BackendApplication.java

@@ -2,8 +2,9 @@
 
 import org.springframework.boot.SpringApplication;
 import org.springframework.boot.autoconfigure.SpringBootApplication;
+import org.springframework.boot.autoconfigure.jdbc.DataSourceAutoConfiguration;
 
-@SpringBootApplication
+@SpringBootApplication(exclude = DataSourceAutoConfiguration.class)
 public class BackendApplication {
     public static void main(String[] args) {
         SpringApplication.run(BackendApplication.class, args);

backend/src/main/java/org/capstone2024/onlyu/controller/FlaskController.java

@@ -0,0 +1,45 @@
+package org.capstone2024.onlyu.controller;
+
+import com.google.gson.JsonObject;
+import lombok.extern.slf4j.Slf4j;
+import org.capstone2024.onlyu.service.FlaskService;
+import org.springframework.beans.factory.annotation.Autowired;
+import org.springframework.http.*;
+import org.springframework.web.bind.annotation.RequestMapping;
+import org.springframework.web.bind.annotation.RequestMethod;
+import org.springframework.web.bind.annotation.RequestParam;
+import org.springframework.web.bind.annotation.RestController;
+import org.springframework.web.client.RestTemplate;
+import org.springframework.web.multipart.MultipartFile;
+
+@RestController
+@Slf4j
+public class FlaskController {
+    private final FlaskService flaskService;
+    @Autowired
+    public FlaskController(FlaskService flaskService){
+        this.flaskService = flaskService;
+    }
+
+    // 퍼스널 컬러 예측과 피부형 검출
+    @RequestMapping(value = "/start", method = RequestMethod.POST)
+    public String start(@RequestParam("email") String email, @RequestParam("gender") String gender,
+                                @RequestParam("image") MultipartFile multipartFile){
+        String predict_color_res = flaskService.predict_color_flask(multipartFile);
+        String predict_shape_res = flaskService.predict_shape_flask();
+
+        JsonObject obj = new JsonObject();
+        obj.addProperty("predictColorRes", predict_color_res);
+        obj.addProperty("predictShapeRes", predict_shape_res);
+        return obj.toString();
+    }
+
+//    @RequestMapping(value = "/gan_image", method = RequestMethod.GET)
+//    public String gan_image(){
+        // Boolean res = flaskService.gan_image_flask();
+//        JsonObject obj = new JsonObject();
+//        obj.addProperty("success", res);
+//        return obj.toString();
+//    }
+
+}

backend/src/main/java/org/capstone2024/onlyu/controller/TestController.java

@@ -3,13 +3,20 @@
 import org.springframework.stereotype.Controller;
 import org.springframework.ui.Model;
 import org.springframework.web.bind.annotation.GetMapping;
+import org.springframework.web.bind.annotation.PostMapping;
 import org.springframework.web.bind.annotation.ResponseBody;
 
 @Controller
 public class TestController {
-    @GetMapping("/")
+    @PostMapping("/")
     @ResponseBody
     public String test(Model model){
+        // 이미지, 성별, 생성형 이미지 유무, 이메일
+        // flask api인 post 요청으로 퍼스널 컬러 검사 결과 요청
+        // get요청으로 predict_shape()해서 얼굴형 받기
         return "Only You~~";
     }
+
+    // 하나 더 만들어서 생성형 이미지 유무에 맞게 요청 보내기
+    // 요거는 front에서 구분해서 보내주면 됨. 대신 -> 로딩 화면 X
 }

backend/src/main/java/org/capstone2024/onlyu/service/FlaskService.java

@@ -0,0 +1,41 @@
+package org.capstone2024.onlyu.service;
+
+import lombok.RequiredArgsConstructor;
+import lombok.extern.slf4j.Slf4j;
+import org.springframework.http.HttpEntity;
+import org.springframework.http.HttpHeaders;
+import org.springframework.http.MediaType;
+import org.springframework.stereotype.Service;
+import org.springframework.transaction.annotation.Transactional;
+import org.springframework.util.LinkedMultiValueMap;
+import org.springframework.util.MultiValueMap;
+import org.springframework.web.client.RestTemplate;
+import org.springframework.web.multipart.MultipartFile;
+
+
+@Service
+@RequiredArgsConstructor
+@Slf4j
+public class FlaskService {
+
+    @Transactional
+    public String predict_color_flask(MultipartFile image){
+        RestTemplate restTemplate = new RestTemplate();
+        HttpHeaders headers = new HttpHeaders();
+        headers.setContentType(MediaType.MULTIPART_FORM_DATA);
+        String url = "http://127.0.0.1:5050/predict_color";
+
+        MultiValueMap<String, Object> body = new LinkedMultiValueMap<>();
+        body.add("image", image.getResource());
+        HttpEntity<MultiValueMap<String, Object>> requestEntity = new HttpEntity<>(body, headers);
+        return restTemplate.postForObject(url, requestEntity, String.class);
+    }
+
+    @Transactional
+    public String predict_shape_flask(){
+        RestTemplate restTemplate = new RestTemplate();
+        String url = "http://127.0.0.1:5050/predict_shape";
+
+        return restTemplate.getForObject(url, String.class);
+    }
+}

main.py

@@ -3,8 +3,6 @@
 from PC_model.pc_model import PersonalColorModel
 from image_processing.gamma_correction import gamma_correction
 
-from sklearn.preprocessing import StandardScaler
-
 import joblib
 import os
 import pandas as pd
@@ -33,15 +31,22 @@ def predict_color():
     global features, pc_model, ss, current_image_path
     # 이미지 저장
     f = request.files['image']
-    f_path = os.path.join(image_path, f.filename)
-    filename = f.filename
+
+    type = f.filename[f.filename.rfind("."):]
+
+    folder_path = os.path.join(image_path, f.filename.replace(".", "_"))
+    f_path = os.path.join(folder_path, "origin_img" + type)
+
+    if not os.path.exists(folder_path):
+        os.makedirs(folder_path)
+
     if not os.path.exists(f_path):
         f.save(f_path)
 
     current_image_path = f_path[:]
 
     # 데이터 추출
-    data = total_data_extract(f_path)
+    data = total_data_extract(f_path, True)
 
     # DataFrame으로 변환
     df = pd.DataFrame(data, index = [0])
@@ -54,6 +59,7 @@ def predict_color():
 
     # 예츨 결과
     raw_res = pc_model.test(preprocssing_data)
+
 
     predict_res = [""] * len(raw_res)
     for idx, predict in enumerate(raw_res):
@@ -75,7 +81,7 @@ def test():
 from shape_detect.controller import get_shape
 @app.route('/predict_shape', methods =['GET'])
 def predict_shape():
-    global image_path
+    global current_image_path
     result = get_shape(current_image_path)
 
     if result == -1: