Add files via upload

Added all the files.

Author: codewithpandey

__pycache__/app.cpython-39.pyc

@@ -0,0 +1,121 @@
+import tkinter as tk
+from tkinter import simpledialog
+import cv2 as cv
+import os
+import PIL.Image, PIL.ImageTk
+import camera
+import model
+import sys
+
+sys.setrecursionlimit(3000)
+
+class App:
+
+    def __init__(self, window = tk.Tk(), window_title = "Recognizer"):
+
+        self.window = window
+        self.window_title = window_title
+
+        self.counters = [1, 1]
+
+        self.model = model.Model()
+
+        self.autoPredict = False
+
+        self.camera = camera.Camera()
+
+        self.startGUI()
+
+        self.delay = 15
+        self.update()
+
+        self.window.attributes('-topmost', True)
+        self.window.mainloop()
+
+    def startGUI(self):
+        self.canvas = tk.Canvas(self.window, width=self.camera.width, height=self.camera.height)
+        self.canvas.pack(anchor=tk.CENTER, expand=True)
+
+        self.btn_toggleauto = tk.Button(self.window, text="Auto Prediction", width=50, command=self.autoPredictToggle)
+        self.btn_toggleauto.pack(anchor=tk.CENTER, expand=True)
+
+        self.classname_one = simpledialog.askstring("Classname One", "Enter the name of the first class: ", parent=self.window)
+        self.classname_two = simpledialog.askstring("Classname Two", "Enter the name of the second class: ", parent=self.window)
+
+        self.btn_class_one = tk.Button(self.window, text=self.classname_one, width=50, command=lambda: self.saveForClass(1))
+        self.btn_class_one.pack(anchor=tk.CENTER, expand=True)
+
+        self.btn_class_two = tk.Button(self.window, text=self.classname_two, width=50, command=lambda: self.saveForClass(2))
+        self.btn_class_two.pack(anchor=tk.CENTER, expand=True)
+
+        self.btn_train = tk.Button(self.window, text="Train Model", width=50, command=lambda: self.model.trainModel(self.counters))
+        self.btn_train.pack(anchor=tk.CENTER, expand=True)
+
+        self.btn_predict = tk.Button(self.window, text="Predict", width=50, command=self.predict)
+        self.btn_predict.pack(anchor=tk.CENTER, expand=True)
+
+        self.btn_reset = tk.Button(self.window, text="Reset", width=50, command=self.reset)
+        self.btn_reset.pack(anchor=tk.CENTER, expand=True)
+
+        self.class_label = tk.Label(self.window, text="CLASS")
+        self.class_label.config(font=("Calibri", 20))
+        self.class_label.pack(anchor=tk.CENTER, expand=True)
+
+
+    def autoPredictToggle(self):
+        self.autoPredict = not self.autoPredict
+
+    def saveForClass(self, class_num):
+        ret, frame = self.camera.getFrame()
+        if not os.path.exists('class1'):
+            os.mkdir('class1')
+        if not os.path.exists('class2'):
+            os.mkdir('class2')
+
+        cv.imwrite(f'class{class_num}/frame{self.counters[class_num - 1]}.jpg', cv.cvtColor(frame, cv.COLOR_RGB2GRAY))
+        img = PIL.Image.open(f'class{class_num}/frame{self.counters[class_num - 1]}.jpg')
+        img.thumbnail((150, 150), PIL.Image.ANTIALIAS)
+        img.save(f'class{class_num}/frame{self.counters[class_num - 1]}.jpg')
+
+        self.counters[class_num - 1] += 1
+
+    
+    def reset(self):
+        for directory in ['class1', 'class2']:
+            for image in os.listdir(directory):
+                file_path = os.path.join(directory, image)
+                if os.path.isfile(file_path):
+                    os.unlink(file_path)
+
+        self.counters = [1, 1]
+        self.model = model.Model()
+        self.class_label.config(text="CLASS")
+
+
+    def update(self):
+        if self.autoPredict:
+            self.predict()
+            pass
+
+        ret, frame = self.camera.getFrame()
+
+        if ret:
+            self.photo = PIL.ImageTk.PhotoImage(image=PIL.Image.fromarray(frame))
+            self.canvas.create_image(0, 0, image=self.photo, anchor=tk.NW)
+
+        print("Updating...")
+        self.window.after(self.delay, self.update)
+
+    
+    def predict(self):            
+        frame = self.camera.getFrame()
+        prediction = self.model.predict(frame)
+
+        if prediction == 1:
+            self.class_label.config(text=self.classname_one)
+            return self.classname_one
+
+        if prediction == 2:
+            self.class_label.config(text=self.classname_two)
+            return self.classname_two
+

__pycache__/camera.cpython-39.pyc

@@ -0,0 +1,42 @@
+from sklearn.svm import LinearSVC
+import numpy as np
+import cv2 as cv
+import PIL
+
+class Model:
+
+    def __init__(self):
+        self.model = LinearSVC()
+
+    def trainModel(self):
+        img_list = np.array([])
+        class_list = np.array([])
+
+        for i in range(1, counters[0]):
+            img = cv.imread(f'class1/frame{i}.jpg')[:,:,0]
+            img = img.reshape(16800) # product of your resolutions
+            img_list = np.append(img_list, [img])
+            class_list = np.append(class_list, 1)
+        
+        for i in range(1, counters[0]):
+            img = cv.imread(f'class2/frame{i}.jpg')[:,:,0]
+            img = img.reshape(16800) # product of your resolutions
+            img_list = np.append(img_list, [img])
+            class_list = np.append(class_list, 2)
+
+        img_list = img_list.reshape(counters[0] - 1 + counters[1] - 1, 16800)
+        self.model.fit(img_list, class_list)
+        print("Model Successfully Trained")
+
+    def predict(self, frame):
+        frame = frame[1]
+        cv.imwrite('frame.jpg', cv.cvtColor(frame, cv.COLOR_RGB2GRAY))
+        img = PIL.Image.open('frame.jpg')
+        img.thumbnail((150, 150), PIL.Image.ANTIALIAS)
+        img.save('frame.jpg')
+
+        img = cv.imread(frame.jpg)[:,:,0]
+        img = img.reshape(16800)
+        prediction = self.model.predict([img])
+
+        return prediction[0]

__pycache__/model.cpython-39.pyc

@@ -0,0 +1,29 @@
+import cv2 as cv
+
+class Camera:
+
+    def __init__(self):
+        self.camera = cv.VideoCapture(0)
+        if not self.camera.isOpened():
+            raise ValueError("Unable to open the camera!")
+
+        self.width = self.camera.get(cv.CAP_PROP_FRAME_WIDTH)
+        self.height = self.camera.get(cv.CAP_PROP_FRAME_HEIGHT)
+
+
+    def __del__(self):
+        if self.camera.isOpened():
+            self.camera.release()
+
+
+    def getFrame(self):
+        if self.camera.isOpened():
+            ret, frame = self.camera.read()
+
+            if ret:
+                return (ret, cv.cvtColor(frame, cv.COLOR_BGR2RGB))
+            else:
+                return (ret, None)
+        else:
+            return None
+        

app.py

@@ -0,0 +1,10 @@
+import app
+import sys
+
+print(sys.getrecursionlimit())
+
+def main():
+    app.App(window_title = "Camera Classifier")
+
+if __name__ == '__main__':
+    main()

brain.py

@@ -0,0 +1,43 @@
+from sklearn.svm import LinearSVC
+import numpy as np
+import cv2 as cv
+import PIL
+# from tensorflow.keras import layers, models
+
+class Model:
+
+    def __init__(self):
+        self.model = LinearSVC()
+
+    def trainModel(self, counters):
+        img_list = np.array([])
+        class_list = np.array([])
+
+        for i in range(1, counters[0]):
+            img = cv.imread(f'class1/frame{i}.jpg')[:,:,0]
+            img = img.reshape(150, 113) # product of your resolutions
+            img_list = np.append(img_list, [img])
+            class_list = np.append(class_list, 1)
+        
+        for j in range(1, counters[0]):
+            img = cv.imread(f'class2/frame{j}.jpg')[:,:,0]
+            img = img.reshape(150, 113) # product of your resolutions
+            img_list = np.append(img_list, [img])
+            class_list = np.append(class_list, 2)
+
+        img_list = img_list.reshape(counters[0] - 1 + counters[1] - 1, 16950)
+        self.model.fit(img_list, class_list)
+        print("Model Successfully Trained")
+
+    def predict(self, frame):
+        frame = frame[1]
+        cv.imwrite('frame.jpg', cv.cvtColor(frame, cv.COLOR_RGB2GRAY))
+        img = PIL.Image.open('frame.jpg')
+        img.thumbnail((150, 150), PIL.Image.ANTIALIAS)
+        img.save('frame.jpg')
+
+        img = cv.imread('frame.jpg')[:,:,0]
+        img = img.reshape(16950)
+        prediction = self.model.predict([img])
+
+        return prediction[0]