在pyplot中添加图像会降低其分辨率吗? [英] Does adding images in pyplot lowers their resolution?

问题描述

以下代码

from PyQt5.QtWidgets import *
from PyQt5.QtCore import *
from PyQt5.QtGui import *

import sys

from matplotlib.backends.backend_qt5agg import FigureCanvasQTAgg as FigureCanvas
from matplotlib.figure import Figure

from matplotlib.offsetbox import OffsetImage, AnnotationBbox

import matplotlib.pyplot as plt
import numpy as np

class View(QGraphicsView):

    def __init__(self):
        super(View, self).__init__()

        self.initScene(5)

    def initScene(self,h):     

        self.scene = QGraphicsScene()
        self.figure = plt.figure()
        self.canvas = FigureCanvas(self.figure)
        self.figure.subplots_adjust(left=0.03,right=1,bottom=.1,top=1,wspace=0, hspace=0)

        ax = self.figure.add_subplot(111)
        ax.set_xlim([0,1000])
        data = np.random.rand(1000)
        ax.plot(data, '-') 

        arr_img = plt.imread('sampleimage.jpg',format='jpg')
        im = OffsetImage(arr_img,zoom=.9)

        ab = AnnotationBbox(im, (.5, .5), xycoords='axes fraction')
        ax.add_artist(ab)

        self.canvas.draw()
        self.setScene(self.scene)
        self.scene.addWidget(self.canvas)

class MainWindow(QMainWindow):

    def __init__(self):
        super(MainWindow,self).__init__()

        #self.setGeometry(150, 150, 700, 550) 

        self.view = View()
        self.setCentralWidget(self.view)

app = QApplication(sys.argv)

window = MainWindow()
window.show()

app.exec_()

产生左下方的输出.右边是我在代码中导入的原始图像("sampleimage.jpg").

produces the output seen below on the left. On the right, is the original image ('sampleimage.jpg') which I imported in the code.

分辨率上的差异很明显.有没有办法在保持质量的前提下将图像添加到地块?

The difference in resolution is apparent. Is there a way to add images to plots, whilst retaining their quality?

推荐答案

在问题代码中，OffsetImage被赋予了参数zoom=0.9.这意味着原始图像的每个像素在屏幕上占0.9/0.72 = 1.25像素.因此，原始图像的5个像素需要在屏幕上压缩为4个像素.这不可避免地会导致在代码输出中观察到一些伪像.

In the code from the question the OffsetImage is given an argument zoom=0.9. This means that each pixel of the original image takes 0.9/0.72=1.25 pixels on screen. Hence 5 pixels of the original image needs to squeezed into 4 pixels on screen. This inevitably leads to some artifacts as observed in the output of the code.

如果要求以原始图像的精确分辨率显示图像，则需要确保OffsetImage每像素恰好使用一个像素.这可以通过将缩放比例设置为ppi为72.除以数字dpi(默认为100)来实现.

If the requirement is to show the image in the exact resolution of the original image, you need to make sure to use exactly one pixel per pixel for the OffsetImage. This would be accomplished by setting the zoom to the ppi of 72. divided by the figure dpi (100 by default).

OffsetImage(arr_img, zoom=72./self.figure.dpi)

结果，显示的图像在matplotlib图中确实具有与原始图像相同的尺寸.

As a result, the image shown would indeed have the same dimensions in the matplotlib plot as the original image.

这篇关于在pyplot中添加图像会降低其分辨率吗?的文章就介绍到这了，希望我们推荐的答案对大家有所帮助，也希望大家多多支持IT屋！