Digit Detector App 1.1

import cv2
import pytesseract
import os
import re
import time
import joblib
import numpy as np
from datetime import datetime
import tkinter as tk
from tkinter import ttk
from PIL import Image, ImageTk
from sklearn.ensemble import RandomForestClassifier

# Constants
IMG_SIZE = (32, 32)
COOLDOWN = 5  # seconds
MIN_OCR_CONF = 60
MIN_ML_CONF = 50
ACTIVE_LEARNING_BATCH = 10
SCAN_INTERVAL = 2.0  # seconds between retries
MIN_WHITE_PIXELS = 500  # skip low-content frames
DATASET_DIR = "dataset"
MODEL_PATH = "model.pkl"

class DigitDetectorApp:
    def __init__(self, model_path=MODEL_PATH):
        # Load or initialize model
        if os.path.exists(model_path):
            self.model = joblib.load(model_path)
        else:
            self.model = RandomForestClassifier(n_estimators=100)
        self.new_data = []

        # Tesseract config
        pytesseract.pytesseract.tesseract_cmd = r"C:\\Program Files\\Tesseract-OCR\\tesseract.exe"
        self.ocr_config = r'--oem 3 --psm 6 -c tessedit_char_whitelist=0123456789'

        # Output dirs
        os.makedirs("detections", exist_ok=True)
        os.makedirs(DATASET_DIR, exist_ok=True)
        self.feedback_file = "feedback_log.csv"
        if not os.path.exists(self.feedback_file):
            with open(self.feedback_file, "w") as f:
                f.write("number,confidence,label,timestamp\n")

        # State
        self.pending_feedback = False
        self.freeze_mode = False
        self.manual_mode = False
        self.last_detected = None
        self.last_time = 0
        self.last_scan_time = 0
        self.history = []
        self.current = None
        self.current_frame_raw = None
        self.current_thresh = None

        # Setup GUI
        self.root = tk.Tk()
        self.root.title("Digit Detector App")
        self.root.geometry("1280x650")
        self._build_gui()

        # Video capture
        self.cap = cv2.VideoCapture(0)
        self.update_frames()
        self.root.mainloop()

    def _build_gui(self):
        frame = self.root
        frame.grid_columnconfigure(0, weight=1)
        frame.grid_columnconfigure(1, weight=1)
        frame.grid_rowconfigure(0, weight=1)

        self.label_cam = ttk.Label(frame)
        self.label_cam.grid(row=0, column=0, padx=5, pady=5, sticky="nsew")
        self.label_proc = ttk.Label(frame)
        self.label_proc.grid(row=0, column=1, padx=5, pady=5, sticky="nsew")

        self.label_info = ttk.Label(frame, text="🧠 Waiting for valid frame...", font=('Arial', 14))
        self.label_info.grid(row=1, column=0, columnspan=2)

        btn_frame = ttk.Frame(frame)
        btn_frame.grid(row=2, column=0, columnspan=2)
        ttk.Button(btn_frame, text="✅ Correct", command=lambda: self.save_feedback(True)).pack(side="left", padx=5)
        ttk.Button(btn_frame, text="❌ Wrong", command=lambda: self.save_feedback(False)).pack(side="left", padx=5)
        ttk.Button(btn_frame, text="🛑 Cancel Freeze", command=self.cancel_freeze).pack(side="left", padx=5)
        ttk.Button(btn_frame, text="🔁 Retry", command=self.force_retry).pack(side="left", padx=5)
        ttk.Button(btn_frame, text="⟲ Undo", command=self.undo).pack(side="left", padx=5)

        self.entry = ttk.Entry(frame)
        self.btn_submit = ttk.Button(frame, text="Submit", command=lambda: self.manual_correction(self.entry.get()))
        self.entry.grid(row=3, column=0, padx=5, pady=5)
        self.btn_submit.grid(row=3, column=1, padx=5, pady=5)
        self.entry.grid_remove(); self.btn_submit.grid_remove()

    def cancel_freeze(self):
        self.freeze_mode = False
        self.pending_feedback = False
        self.manual_mode = False
        self.entry.grid_remove(); self.btn_submit.grid_remove()
        self.label_info.config(text="⏹️ Freeze canceled. Resuming live detection.")

    def force_retry(self):
        self.last_scan_time = 0
        self.label_info.config(text="🔁 Retry triggered.")

    def preprocess(self, frame):
        gray = cv2.cvtColor(frame, cv2.COLOR_BGR2GRAY)
        blur = cv2.GaussianBlur(gray, (3,3),0)
        _, thresh = cv2.threshold(blur,170,255,cv2.THRESH_BINARY)
        return thresh

    def predict_ocr(self, img):
        data = pytesseract.image_to_data(img, config=self.ocr_config, output_type=pytesseract.Output.DICT)
        for i, txt in enumerate(data['text']):
            txt = txt.strip()
            try: conf = int(data['conf'][i])
            except: conf = 0
            if re.fullmatch(r'\d{1,3}', txt) and conf >= MIN_OCR_CONF:
                return txt, conf
        return None, 0

    def predict_ml(self, img):
        resized = cv2.resize(img, IMG_SIZE).flatten().reshape(1,-1)
        pred = self.model.predict(resized)[0]
        conf = int(max(self.model.predict_proba(resized)[0]) *100) if hasattr(self.model,'predict_proba') else 0
        return str(pred), conf

    def ensemble_predict(self, img):
        # Only use OCR; disable ML fallback to avoid false positives on empty or noisy frames
        num_o, conf_o = self.predict_ocr(img)
        # Accept only OCR predictions with sufficient confidence
        if num_o and conf_o >= MIN_OCR_CONF:
            return num_o, conf_o
        return None, 0

    def save_feedback(self, correct):
        if not self.current: return
        num, conf, thresh = self.current
        ts = datetime.now().strftime('%Y-%m-%d_%H-%M-%S')
        if correct:
            # log and unfreeze
            with open(self.feedback_file,'a') as f:
                f.write(f"{num},{conf},correct,{ts}\n")
            folder=os.path.join(DATASET_DIR,num)
            os.makedirs(folder,exist_ok=True)
            cv2.imwrite(os.path.join(folder,f"{ts}.png"), self.current_frame_raw)
            self.new_data.append((thresh,int(num)))
            if len(self.new_data)>=ACTIVE_LEARNING_BATCH: self.retrain_model()
            self.freeze_mode=False; self.pending_feedback=False; self.manual_mode=False
            self.entry.grid_remove(); self.btn_submit.grid_remove()
            self.label_info.config(text=f"✅ Saved correct: {num}")
        else:
            # freeze and enter manual mode
            self.freeze_mode=True; self.pending_feedback=True; self.manual_mode=False
            self.label_info.config(text=f"❌ Wrong. Will retry or await correction.")

    def manual_correction(self,val):
        if not self.current: return
        val=val.strip()
        if not(val.isdigit() or val.lower()=='nothing'):
            self.label_info.config(text="⚠️ Enter digit or 'nothing'")
            return
        num_old,conf_old,thresh=self.current
        ts=datetime.now().strftime('%Y-%m-%d_%H-%M-%S')
        folder=val if val.isdigit() else 'unlabeled'
        path=os.path.join(DATASET_DIR,folder)
        os.makedirs(path,exist_ok=True)
        cv2.imwrite(os.path.join(path,f"{ts}.png"),self.current_frame_raw)
        with open(self.feedback_file,'a') as f:
            f.write(f"{val},{conf_old},manual,{ts}\n")
        # update
        self.current=(val,conf_old,self.current_thresh)
        self.freeze_mode=False; self.pending_feedback=False; self.manual_mode=False
        self.entry.grid_remove(); self.btn_submit.grid_remove()
        self.label_info.config(text=f"📤 Correction saved: {val}")

    def undo(self):
        if not self.history: return
        num,ts,thresh=self.history.pop()
        self.label_info.config(text=f"⟲ Undone: {num} at {ts}")
        self.display_images(thresh)

    def retrain_model(self):
        X,y=[],[]
        for label in os.listdir(DATASET_DIR):
            if label=='unlabeled': continue
            for fn in os.listdir(os.path.join(DATASET_DIR,label)):
                img=cv2.imread(os.path.join(DATASET_DIR,label,fn),cv2.IMREAD_GRAYSCALE)
                if img is None: continue
                X.append(cv2.resize(img,IMG_SIZE).flatten()); y.append(int(label))
        for img,lbl in self.new_data:
            X.append(img.flatten()); y.append(lbl)
        if X:
            self.model=RandomForestClassifier(n_estimators=100)
            self.model.fit(np.array(X),np.array(y))
            joblib.dump(self.model,MODEL_PATH)
            self.new_data.clear()
            self.label_info.config(text="🤖 Model retrained.")

    def display_images(self,proc):
        frame=cv2.resize(self.last_frame,(600,450))
        cam=ImageTk.PhotoImage(Image.fromarray(cv2.cvtColor(frame,cv2.COLOR_BGR2RGB)))
        pr=ImageTk.PhotoImage(Image.fromarray(proc))
        self.label_cam.config(image=cam); self.label_cam.imgtk=cam
        self.label_proc.config(image=pr); self.label_proc.imgtk=pr

    def update_frames(self):
        now=time.time()
        # Freeze & retry
        if self.freeze_mode and self.current and not self.manual_mode:
            if now-self.last_scan_time>SCAN_INTERVAL:
                # retry detection on saved thresh
                num,conf=self.ensemble_predict(self.current_thresh)
                self.current=(num,conf,self.current_thresh)
                self.label_info.config(text=f"🔁 Retry: Was {num} correct?")
                self.last_scan_time=now
                # after first retry allow manual
                self.manual_mode=True
                self.entry.grid(); self.btn_submit.grid()
            self.display_images(self.current[2])
            self.root.after(30,self.update_frames)
            return

        # live capture
        ret,frame=self.cap.read()
        if not ret:
            self.root.after(30,self.update_frames); return
        self.last_frame=frame
        proc=self.preprocess(frame)
        self.current_frame_raw=frame.copy()
        self.current_thresh=proc

        # skip low-content
        if cv2.countNonZero(proc)<MIN_WHITE_PIXELS:
            self.label_info.config(text="📷 Weak image, skipping.")
            self.display_images(proc)
            self.root.after(30,self.update_frames)
            return

        # initial detection
        if not self.pending_feedback:
            # validate detection
            num_o,conf_o=self.predict_ocr(proc)
            num_m,conf_m=self.predict_ml(proc)
            valid_o=num_o and conf_o>=MIN_OCR_CONF
            valid_m=num_m and conf_m>=MIN_ML_CONF
            if not(valid_o or valid_m):
                self.root.after(30,self.update_frames); return
            num,conf=(num_o,conf_o) if valid_o and conf_o>=conf_m else (num_m,conf_m)
            # freeze and prompt
            ts=datetime.now().strftime('%Y-%m-%d_%H-%M-%S')
            cv2.imwrite(f"detections/detected_{ts}.png",frame)
            self.history.append((num,ts,proc))
            self.current=(num,conf,proc)
            self.last_detected, self.last_time = num, now
            self.pending_feedback=True; self.freeze_mode=True; self.manual_mode=False
            self.label_info.config(text=f"❓ Was {num} correct?")
            self.entry.delete(0,tk.END); self.entry.grid(); self.btn_submit.grid()

        self.display_images(proc)
        self.root.after(30,self.update_frames)

if __name__=="__main__":
    DigitDetectorApp()