【DeeplabV3+ 指标可视化】DeeplabV3+评价指标可视化

1 评价指标可视化用在何处

在文章DeeplabV3+计算评价指标中给出了分割模型的评价指标有哪些，以及如何计算，在得到验证集的各项指标后，没点可视化操作显得不够专业，故增加此文。

if __name__ == "__main__":
        # --------------------------------------------------------#
        #   hist：验证集的混淆矩阵，shape：(21,21)
        #   IoUs：每个类别的IoU，shape：(21,)
        #   PA_Recall：每个类别的像素精度召回率(行和)，shape：(21,)
        #   Precision：每个类别的像素精度(列和)，shape:(21,)
        #	每一行之和是该类的真实样本数量，每一列之和是预测为该类的样本数量。
        # --------------------------------------------------------# 
        hist, IoUs, PA_Recall, Precision = compute_mIoU(gt_dir, pred_dir, image_ids, num_classes, name_classes)  # 执行计算评价指标的函数
        # --------------------------------------------------------# 
        #   miou_out_path：输出结果路径
        # --------------------------------------------------------# 
        show_results(miou_out_path, hist, IoUs, PA_Recall, Precision, name_classes)

2 show_results()代码理解

import csv
import os
from os.path import join

import matplotlib.pyplot as plt
import numpy as np
from PIL import Image


def adjust_axes(r, t, fig, axes):
    bb                  = t.get_window_extent(renderer=r)
    text_width_inches   = bb.width / fig.dpi
    current_fig_width   = fig.get_figwidth()
    new_fig_width       = current_fig_width + text_width_inches
    propotion           = new_fig_width / current_fig_width
    x_lim               = axes.get_xlim()
    axes.set_xlim([x_lim[0], x_lim[1] * propotion])

def draw_plot_func(values, name_classes, plot_title, x_label, output_path, tick_font_size = 12, plt_show = True):
    """
    values: 一维列表
    name_classes: 种类名称，例如这样：["_background_","cat","dog",...]
    """
    fig     = plt.gcf()     # 获得当前图表
    axes    = plt.gca()     # 获得当前子图
    # --------------------------------------------#
    #   plt.bar：   正常纵向柱状图
    #   plt.barh：  横向柱状图
    #       参数：纵坐标有哪些，每个有多宽(高)，颜色
    # --------------------------------------------#
    plt.barh(range(len(values)), values, color='royalblue')
    plt.title(plot_title, fontsize=tick_font_size + 2)
    plt.xlabel(x_label, fontsize=tick_font_size)
    #   name_classes：种类名称，例如这样：["_background_","cat","dog",...]
    plt.yticks(range(len(values)), name_classes, fontsize=tick_font_size)
    r = fig.canvas.get_renderer()       # <matplotlib.backends.backend_agg.RendererAgg object at 0x00000244FFFFA550>
    for i, val in enumerate(values):
        str_val = " " + str(val) 
        if val < 1.0:
            str_val = " {0:.2f}".format(val)
        # -----------------------------------------------------#
        #   plt.text：用于设置文字说明
        #       参数：前两个参数组成说明文字的横纵坐标，三参：说明文字
        # -----------------------------------------------------#
        t = plt.text(val, i, str_val, color='royalblue', va='center', fontweight='bold')
        if i == (len(values)-1):
            adjust_axes(r, t, fig, axes)        # 用于调整坐标参数

    fig.tight_layout()          # 自动调整子图参数，让图显示好看点
    fig.savefig(output_path)
    if plt_show:
        plt.show()
    plt.close()

def show_results(miou_out_path, hist, IoUs, PA_Recall, Precision, name_classes, tick_font_size = 12):
    # -------------------------------------------------------#
    #   IoUs：          每个类别的IoU，shape：(21,)
    #   PA_Recall：     每个类别的像素精度召回率(行和)，shape：(21,)
    #   Precision：     每个类别的像素精度(列和)，shape:(21,)
    #   name_classes：种类名称，例如这样：["_background_","cat","dog",...]
    #   "mIoU = {0:.2f}%".format(np.nanmean(IoUs)*100)：图片标题
    #   "Intersection over Union"：横坐标
    #   os.path.join(miou_out_path, "mIoU.png")：保存路径
    # -------------------------------------------------------#
    draw_plot_func(IoUs, name_classes, "mIoU = {0:.2f}%".format(np.nanmean(IoUs)*100), "Intersection over Union", \
        os.path.join(miou_out_path, "mIoU.png"), tick_font_size = tick_font_size, plt_show = True)
    print("Save mIoU out to " + os.path.join(miou_out_path, "mIoU.png"))

    draw_plot_func(PA_Recall, name_classes, "mPA = {0:.2f}%".format(np.nanmean(PA_Recall)*100), "Pixel Accuracy", \
        os.path.join(miou_out_path, "mPA.png"), tick_font_size = tick_font_size, plt_show = False)
    print("Save mPA out to " + os.path.join(miou_out_path, "mPA.png"))
    
    draw_plot_func(PA_Recall, name_classes, "mRecall = {0:.2f}%".format(np.nanmean(PA_Recall)*100), "Recall", \
        os.path.join(miou_out_path, "Recall.png"), tick_font_size = tick_font_size, plt_show = False)
    print("Save Recall out to " + os.path.join(miou_out_path, "Recall.png"))

    draw_plot_func(Precision, name_classes, "mPrecision = {0:.2f}%".format(np.nanmean(Precision)*100), "Precision", \
        os.path.join(miou_out_path, "Precision.png"), tick_font_size = tick_font_size, plt_show = False)
    print("Save Precision out to " + os.path.join(miou_out_path, "Precision.png"))

    with open(os.path.join(miou_out_path, "confusion_matrix.csv"), 'w', newline='') as f:
        writer          = csv.writer(f)
        writer_list     = []
        writer_list.append([' '] + [str(c) for c in name_classes])      # 往csv文件里写坐标名称
        for i in range(len(hist)):      # 逐行读写混淆矩阵参数
            writer_list.append([name_classes[i]] + [str(x) for x in hist[i]])   # 里面的for逐列
        writer.writerows(writer_list)
    print("Save confusion_matrix out to " + os.path.join(miou_out_path, "confusion_matrix.csv"))

3 输出结果展示

confusion_matrix.csv内容

MIoU.png内容

mPA.png内容

Precision.png内容

Recall.png内容

4 感谢链接

https://blog.csdn.net/weixin_44791964/article/details/120113686