diff --git a/README.md b/README.md
index bd505ded45889a6c3298fc4fb85f580e1ea0fc9e..cd6f5d3400709f1a607521a1a817849113b2deda 100644
--- a/README.md
+++ b/README.md
@@ -84,7 +84,8 @@ Dive into Deep Learning with PyTorch.
### 9. 计算机视觉
[9.1 图像增广](https://github.com/ShusenTang/Dive-into-DL-PyTorch/blob/master/docs/chapter09_computer-vision/9.1_image-augmentation.md)
-[9.2 微调](https://github.com/ShusenTang/Dive-into-DL-PyTorch/blob/master/docs/chapter09_computer-vision/9.2_fine-tuning.md)
+[9.2 微调](https://github.com/ShusenTang/Dive-into-DL-PyTorch/blob/master/docs/chapter09_computer-vision/9.2_fine-tuning.md)
+[9.3 目标检测和边界框](https://github.com/ShusenTang/Dive-into-DL-PyTorch/blob/master/docs/chapter09_computer-vision/9.3_bounding-box.md)
持续更新中......
diff --git a/code/chapter09_computer-vision/9.3_bounding-box.ipynb b/code/chapter09_computer-vision/9.3_bounding-box.ipynb
new file mode 100644
index 0000000000000000000000000000000000000000..599ebb086924c6c7c7a02430093fac4b23093816
--- /dev/null
+++ b/code/chapter09_computer-vision/9.3_bounding-box.ipynb
@@ -0,0 +1,944 @@
+{
+ "cells": [
+ {
+ "cell_type": "markdown",
+ "metadata": {},
+ "source": [
+ "# 9.3 目标检测和边界框"
+ ]
+ },
+ {
+ "cell_type": "code",
+ "execution_count": 1,
+ "metadata": {
+ "collapsed": true
+ },
+ "outputs": [],
+ "source": [
+ "%matplotlib inline\n",
+ "from PIL import Image\n",
+ "\n",
+ "import sys\n",
+ "sys.path.append(\"..\") \n",
+ "import d2lzh_pytorch as d2l"
+ ]
+ },
+ {
+ "cell_type": "code",
+ "execution_count": 2,
+ "metadata": {},
+ "outputs": [
+ {
+ "data": {
+ "image/svg+xml": [
+ "\n",
+ "\n",
+ "\n",
+ "\n"
+ ],
+ "text/plain": [
+ ""
+ ]
+ },
+ "metadata": {},
+ "output_type": "display_data"
+ }
+ ],
+ "source": [
+ "d2l.set_figsize()\n",
+ "img = Image.open('../../img/catdog.jpg')\n",
+ "d2l.plt.imshow(img); # 加分号只显示图"
+ ]
+ },
+ {
+ "cell_type": "markdown",
+ "metadata": {},
+ "source": [
+ "## 9.3.1 边界框"
+ ]
+ },
+ {
+ "cell_type": "code",
+ "execution_count": 3,
+ "metadata": {
+ "collapsed": true
+ },
+ "outputs": [],
+ "source": [
+ "# bbox是bounding box的缩写\n",
+ "dog_bbox, cat_bbox = [60, 45, 378, 516], [400, 112, 655, 493]"
+ ]
+ },
+ {
+ "cell_type": "code",
+ "execution_count": 4,
+ "metadata": {
+ "collapsed": true
+ },
+ "outputs": [],
+ "source": [
+ "def bbox_to_rect(bbox, color): # 本函数已保存在d2lzh_pytorch中方便以后使用\n",
+ " # 将边界框(左上x, 左上y, 右下x, 右下y)格式转换成matplotlib格式:\n",
+ " # ((左上x, 左上y), 宽, 高)\n",
+ " return d2l.plt.Rectangle(\n",
+ " xy=(bbox[0], bbox[1]), width=bbox[2]-bbox[0], height=bbox[3]-bbox[1],\n",
+ " fill=False, edgecolor=color, linewidth=2)"
+ ]
+ },
+ {
+ "cell_type": "code",
+ "execution_count": 5,
+ "metadata": {},
+ "outputs": [
+ {
+ "data": {
+ "image/svg+xml": [
+ "\n",
+ "\n",
+ "\n",
+ "\n"
+ ],
+ "text/plain": [
+ ""
+ ]
+ },
+ "metadata": {},
+ "output_type": "display_data"
+ }
+ ],
+ "source": [
+ "fig = d2l.plt.imshow(img)\n",
+ "fig.axes.add_patch(bbox_to_rect(dog_bbox, 'blue'))\n",
+ "fig.axes.add_patch(bbox_to_rect(cat_bbox, 'red'));"
+ ]
+ },
+ {
+ "cell_type": "code",
+ "execution_count": null,
+ "metadata": {
+ "collapsed": true
+ },
+ "outputs": [],
+ "source": []
+ }
+ ],
+ "metadata": {
+ "kernelspec": {
+ "display_name": "Python [conda env:anaconda3]",
+ "language": "python",
+ "name": "conda-env-anaconda3-py"
+ },
+ "language_info": {
+ "codemirror_mode": {
+ "name": "ipython",
+ "version": 3
+ },
+ "file_extension": ".py",
+ "mimetype": "text/x-python",
+ "name": "python",
+ "nbconvert_exporter": "python",
+ "pygments_lexer": "ipython3",
+ "version": "3.6.8"
+ }
+ },
+ "nbformat": 4,
+ "nbformat_minor": 2
+}
diff --git a/code/d2lzh_pytorch/utils.py b/code/d2lzh_pytorch/utils.py
index 27876f5ee24205df57925357a75bf0947c432135..f4d1bbf74b49fcddad9422a9480f20ce81ab12f4 100644
--- a/code/d2lzh_pytorch/utils.py
+++ b/code/d2lzh_pytorch/utils.py
@@ -690,7 +690,6 @@ def show_images(imgs, num_rows, num_cols, scale=2):
axes[i][j].axes.get_yaxis().set_visible(False)
return axes
-# 本函数已保存在d2lzh_pytorch包中方便以后使用
def train(train_iter, test_iter, net, loss, optimizer, device, num_epochs):
net = net.to(device)
print("training on ", device)
@@ -712,3 +711,15 @@ def train(train_iter, test_iter, net, loss, optimizer, device, num_epochs):
test_acc = evaluate_accuracy(test_iter, net)
print('epoch %d, loss %.4f, train acc %.3f, test acc %.3f, time %.1f sec'
% (epoch + 1, train_l_sum / batch_count, train_acc_sum / n, test_acc, time.time() - start))
+
+
+
+
+
+############################## 9.3 #####################
+def bbox_to_rect(bbox, color):
+ # 将边界框(左上x, 左上y, 右下x, 右下y)格式转换成matplotlib格式:
+ # ((左上x, 左上y), 宽, 高)
+ return d2l.plt.Rectangle(
+ xy=(bbox[0], bbox[1]), width=bbox[2]-bbox[0], height=bbox[3]-bbox[1],
+ fill=False, edgecolor=color, linewidth=2)
\ No newline at end of file
diff --git a/docs/chapter09_computer-vision/9.3_bounding-box.md b/docs/chapter09_computer-vision/9.3_bounding-box.md
new file mode 100644
index 0000000000000000000000000000000000000000..8511466eed0aa5272199d217a957f4e36eb030f3
--- /dev/null
+++ b/docs/chapter09_computer-vision/9.3_bounding-box.md
@@ -0,0 +1,67 @@
+# 9.3 目标检测和边界框
+
+在前面的一些章节中,我们介绍了诸多用于图像分类的模型。在图像分类任务里,我们假设图像里只有一个主体目标,并关注如何识别该目标的类别。然而,很多时候图像里有多个我们感兴趣的目标,我们不仅想知道它们的类别,还想得到它们在图像中的具体位置。在计算机视觉里,我们将这类任务称为目标检测(object detection)或物体检测。
+
+目标检测在多个领域中被广泛使用。例如,在无人驾驶里,我们需要通过识别拍摄到的视频图像里的车辆、行人、道路和障碍的位置来规划行进线路。机器人也常通过该任务来检测感兴趣的目标。安防领域则需要检测异常目标,如歹徒或者炸弹。
+
+在接下来的几节里,我们将介绍目标检测里的多个深度学习模型。在此之前,让我们来了解目标位置这个概念。先导入实验所需的包或模块。
+
+``` python
+%matplotlib inline
+from PIL import Image
+
+import sys
+sys.path.append("..")
+import d2lzh_pytorch as d2l
+```
+
+下面加载本节将使用的示例图像。可以看到图像左边是一只狗,右边是一只猫。它们是这张图像里的两个主要目标。
+
+``` python
+d2l.set_figsize()
+img = Image.open('../../img/catdog.jpg')
+d2l.plt.imshow(img); # 加分号只显示图
+```
+
+
+
+
+## 9.3.1 边界框
+
+在目标检测里,我们通常使用边界框(bounding box)来描述目标位置。边界框是一个矩形框,可以由矩形左上角的$x$和$y$轴坐标与右下角的$x$和$y$轴坐标确定。我们根据上面的图的坐标信息来定义图中狗和猫的边界框。图中的坐标原点在图像的左上角,原点往右和往下分别为$x$轴和$y$轴的正方向。
+
+``` python
+# bbox是bounding box的缩写
+dog_bbox, cat_bbox = [60, 45, 378, 516], [400, 112, 655, 493]
+```
+
+我们可以在图中将边界框画出来,以检查其是否准确。画之前,我们定义一个辅助函数`bbox_to_rect`。它将边界框表示成matplotlib的边界框格式。
+
+``` python
+def bbox_to_rect(bbox, color): # 本函数已保存在d2lzh_pytorch中方便以后使用
+ # 将边界框(左上x, 左上y, 右下x, 右下y)格式转换成matplotlib格式:
+ # ((左上x, 左上y), 宽, 高)
+ return d2l.plt.Rectangle(
+ xy=(bbox[0], bbox[1]), width=bbox[2]-bbox[0], height=bbox[3]-bbox[1],
+ fill=False, edgecolor=color, linewidth=2)
+```
+
+我们将边界框加载在图像上,可以看到目标的主要轮廓基本在框内。
+
+``` python
+fig = d2l.plt.imshow(img)
+fig.axes.add_patch(bbox_to_rect(dog_bbox, 'blue'))
+fig.axes.add_patch(bbox_to_rect(cat_bbox, 'red'));
+```
+输出:
+
+
+
+
+## 小结
+
+* 在目标检测里不仅需要找出图像里面所有感兴趣的目标,而且要知道它们的位置。位置一般由矩形边界框来表示。
+
+
+-----------
+> 注:除代码外本节与原书基本相同,[原书传送门](https://zh.d2l.ai/chapter_computer-vision/bounding-box.html)
diff --git a/img/catdog.jpg b/img/catdog.jpg
new file mode 100644
index 0000000000000000000000000000000000000000..4471a714dc01726531128002517170a0140d4de5
Binary files /dev/null and b/img/catdog.jpg differ
diff --git a/img/chapter09/9.3_output1.png b/img/chapter09/9.3_output1.png
new file mode 100644
index 0000000000000000000000000000000000000000..9a57d44158bafbbd7c5def1bd51be6f7a5cc846e
Binary files /dev/null and b/img/chapter09/9.3_output1.png differ
diff --git a/img/chapter09/9.3_output2.png b/img/chapter09/9.3_output2.png
new file mode 100644
index 0000000000000000000000000000000000000000..9d361b669f356cfc33ee77ca0fa99b16b00ec724
Binary files /dev/null and b/img/chapter09/9.3_output2.png differ