I am trying to use Computer vision to see numbers I've handwritten on a sheet. My code so far is:
import cv2
import numpy as np
import pandas as pd
import seaborn as sns
import matplotlib.pyplot as plt
from sklearn.datasets import fetch_openml
from sklearn.model_selection import train_test_split
from sklearn.linear_model import LogisticRegression
from sklearn.metrics import accuracy_score
from PIL import Image
import PIL.ImageOps
import os, ssl, time
#Setting an HTTPS Context to fetch data from OpenML
if (not os.environ.get('PYTHONHTTPSVERIFY', '') and
getattr(ssl, '_create_unverified_context', None)):
ssl._create_default_https_context = ssl._create_unverified_context
#Fetching the data
X, y = fetch_openml('mnist_784', version=1, return_X_y=True)
print(pd.Series(y).value_counts())
classes = ['0', '1', '2','3', '4','5', '6', '7', '8', '9']
nclasses = len(classes)
#Splitting the data and scaling it
X_train, X_test, y_train, y_test = train_test_split(X, y, random_state=9,
train_size=7500, test_size=2500)
#scaling the features
X_train_scaled = X_train/255.0
X_test_scaled = X_test/255.0
#Fitting the training data into the model
clf = LogisticRegression(solver='saga',multi_class='multinomial').fit(X_train_scaled, y_train)
#Calculating the accuracy of the model
y_pred = clf.predict(X_test_scaled)
accuracy = accuracy_score(y_test, y_pred)
print("The accuracy is :- ",accuracy)
#Starting the camera
cap = cv2.VideoCapture(0)
while(True):
# Capture frame-by-frame
try:
ret, frame = cap.read()
# Our operations on the frame come here
gray = cv2.cvtColor(frame, cv2.COLOR_BGR2GRAY)
#Drawing a box in the center of the video
height, width = gray.shape
upper_left = (int(width / 2 - 56), int(height / 2 - 56))
bottom_right = (int(width / 2 + 56), int(height / 2 + 56))
cv2.rectangle(gray, upper_left, bottom_right, (0, 255, 0), 2)
#To only consider the area inside the box for detecting the digit
#roi = Region Of Interest
roi = gray[upper_left[1]:bottom_right[1], upper_left[0]:bottom_right[0]]
#Converting cv2 image to pil format
im_pil = Image.fromarray(roi)
# convert to grayscale image - 'L' format means each pixel is
# represented by a single value from 0 to 255
image_bw = im_pil.convert('L')
image_bw_resized = image_bw.resize((28,28), Image.ANTIALIAS)
image_bw_resized_inverted = PIL.ImageOps.invert(image_bw_resized)
pixel_filter = 20
min_pixel = np.percentile(image_bw_resized_inverted, pixel_filter)
image_bw_resized_inverted_scaled = np.clip(image_bw_resized_inverted-min_pixel, 0, 255)
max_pixel = np.max(image_bw_resized_inverted)
image_bw_resized_inverted_scaled = np.asarray(image_bw_resized_inverted_scaled)/max_pixel
test_sample = np.array(image_bw_resized_inverted_scaled).reshape(1,784)
test_pred = clf.predict(test_sample)
print("Predicted class is: ", test_pred)
# Display the resulting frame
cv2.imshow('frame',gray)
if cv2.waitKey(1) & 0xFF == ord('q'):
break
except Exception as e:
pass
# When everything is done, release the capture
cap.release()
cv2.destroyAllWindows()
I expected, say,
Predicted class is: ['1']
When I held a handwritten 1 up to the camera. I got:
th feature names Predicted class is: ['3'] Warning (from warnings module): File "C:\Users\RUSHIKESH\OneDrive\Desktop\practice codeds\digits.py", line 66 image_bw_resized = image_bw.resize((28,28), Image.LANCZOS) DeprecationWarning: LANCZOS is deprecated and will be removed in Pillow 10 (2023-07-01). Use Resampling.LANCZOS instead. Warning (from warnings module): File "C:\Users\RUSHIKESH\AppData\Local\Programs\Python\Python310\lib\site-packages\sklearn\base.py", line 450 warnings.warn( UserWarning: X does not have valid feature names, but LogisticRegression was fitted with feature names Predicted class is: ['3'] Warning (from warnings module): File "C:\Users\RUSHIKESH\OneDrive\Desktop\practice codeds\digits.py", line 66 image_bw_resized = image_bw.resize((28,28), Image.LANCZOS) DeprecationWarning: LANCZOS is deprecated and will be removed in Pillow 10 (2023-07-01). Use Resampling.LANCZOS instead. Warning (from warnings module): File "C:\Users\RUSHIKESH\AppData\Local\Programs\Python\Py
No matter what I held up to the camera.
How to fix this?