代码下载：Python+TensorFlow+PyQt实现手写体数字图片识别+GUI界面+画板数字识别

# -*- coding:utf-8 -*- import sys from PyQt5.QtGui import * from PyQt5.QtWidgets import * from PyQt5 import QtCore, QtGui from PIL import Image import tensorflow as tf import matplotlib.pyplot as plt import cv2 import mnist_inference import mnist_train import pathlib class DigitalRecognition(QMainWindow): def __init__(self): super(DigitalRecognition, self).__init__() self.initUI() def initUI(self): self.setWindowTitle("手写数字识别") self.resize(800, 500) _translate = QtCore.QCoreApplication.translate self.button1 = QPushButton("选择图片", self) self.button1.move(150, 350) self.button1.resize(120,60) self.button1.clicked.connect(self.select_image) self.button2 = QPushButton("识别", self) self.button2.move(550, 350) self.button2.resize(120, 60) self.button2.clicked.connect(self.recognition) self.edit = QTextEdit(self) self.edit.move(450, 100) self.edit.resize(300, 160) self.edit.setPlaceholderText(_translate("MainWindow", "结果显示")) self.label = QLabel("待载入图片", self) self.label.move(110, 60) self.label.resize(250, 250) self.label.setStyleSheet("QLabel{background:gray;}" "QLabel{color:rgb(0,0,0,120);font-size:15px;font-weight:bold;font-family:宋体;}") def select_image(self): global fname imgName, imgType = QFileDialog.getOpenFileName(self, "打开图片", "", "*.png;;*.jpg;;All Files(*)") jpg = QtGui.QPixmap(imgName).scaled(self.label.width(), self.label.height()) self.label.setPixmap(jpg) fname = imgName def recognition(self): global fname def imageprepare(): """ This function returns the pixel values. The imput is a png file location. """ file_name = str(fname) # 导入自己的图片地址 im = Image.open(file_name) plt.imshow(im) img = cv2.imread(file_name, 1) ret, th1 = cv2.threshold(img, 127, 255, cv2.THRESH_BINARY) im = Image.fromarray(cv2.cvtColor(th1, cv2.COLOR_BGR2RGB)) # in terminal 'mogrify -format png *.jpg' convert jpg to png #灰度化 im = im.convert('L') # 为图片重新指定尺寸 im = im.resize((28, 28), Image.ANTIALIAS) im_list = list(im.getdata()) # get pixel values # normalize pixels to 0 and 1. 0 is pure white, 1 is pure black. #图像灰度反转 tva = [(255 - x) * 1.0 / 255.0 for x in im_list] # print(tva) return tva # Define the model (same as when creating the model file) # 图像处理：降噪、灰度化、修改尺寸以及灰度反转 pic = imageprepare() def evaluate(pic): with tf.Graph().as_default() as g: x = tf.placeholder(tf.float32, [None, mnist_inference.INPUT_NODE], name='x-input') y_ = tf.placeholder(tf.float32, [None, mnist_inference.OUTPUT_NODE], name='y-input') validate_feed = {x: [pic]} y = mnist_inference.inference(x, None) result = tf.argmax(y, 1) variable_averages = tf.train.ExponentialMovingAverage(mnist_train.MOVING_AVERAGE_DECAY) variables_to_restore = variable_averages.variables_to_restore() saver = tf.train.Saver(variables_to_restore) with tf.Session() as sess: ckpt = tf.train.get_checkpoint_state(mnist_train.MODEL_SAVE_PATH) if ckpt and ckpt.model_checkpoint_path: saver.restore(sess, ckpt.model_checkpoint_path) number = sess.run(result, feed_dict=validate_feed) pic_name = pathlib.Path(fname).name#15元 print(pic_name, ' is :', number[0]) else: print('No checkpoint file found') return self.edit.setText(pic_name+'识别的手写数字是>>>' + str(number[0])) #识别图像 evaluate(pic) #15yuan if __name__ == "__main__": app = QApplication(sys.argv) main = DigitalRecognition() palette = QPalette() pix5 = QPixmap("timg") # 背景图 pix = pix5.scaled(main.width(), main.height()) # 自适应窗口大小 palette.setBrush(QPalette.Background, QBrush(pix)) main.setPalette(palette) main.show() sys.exit(app.exec_())