.. _sec_bbox: Object Detection and Bounding Boxes =================================== In earlier sections (e.g., :numref:`sec_alexnet`–:numref:`sec_googlenet`), we introduced various models for image classification. In image classification tasks, we assume that there is only *one* major object in the image and we only focus on how to recognize its category. However, there are often *multiple* objects in the image of interest. We not only want to know their categories, but also their specific positions in the image. In computer vision, we refer to such tasks as *object detection* (or *object recognition*). Object detection has been widely applied in many fields. For example, self-driving needs to plan traveling routes by detecting the positions of vehicles, pedestrians, roads, and obstacles in the captured video images. Besides, robots may use this technique to detect and localize objects of interest throughout its navigation of an environment. Moreover, security systems may need to detect abnormal objects, such as intruders or bombs. In the next few sections, we will introduce several deep learning methods for object detection. We will begin with an introduction to *positions* (or *locations*) of objects. .. raw:: html
pytorchmxnettensorflow
.. raw:: html
.. raw:: latex \diilbookstyleinputcell .. code:: python %matplotlib inline import torch from d2l import torch as d2l .. raw:: html
.. raw:: html
.. raw:: latex \diilbookstyleinputcell .. code:: python %matplotlib inline from mxnet import image, np, npx from d2l import mxnet as d2l npx.set_np() .. raw:: html
.. raw:: html
.. raw:: latex \diilbookstyleinputcell .. code:: python %matplotlib inline import tensorflow as tf from d2l import tensorflow as d2l .. raw:: html
.. raw:: html
We will load the sample image to be used in this section. We can see that there is a dog on the left side of the image and a cat on the right. They are the two major objects in this image. .. raw:: html
pytorchmxnettensorflow
.. raw:: html
.. raw:: latex \diilbookstyleinputcell .. code:: python d2l.set_figsize() img = d2l.plt.imread('../img/catdog.jpg') d2l.plt.imshow(img); .. figure:: output_bounding-box_d6b70e_15_0.svg .. raw:: html
.. raw:: html
.. raw:: latex \diilbookstyleinputcell .. code:: python d2l.set_figsize() img = image.imread('../img/catdog.jpg').asnumpy() d2l.plt.imshow(img); .. raw:: latex \diilbookstyleoutputcell .. parsed-literal:: :class: output [21:49:41] ../src/storage/storage.cc:196: Using Pooled (Naive) StorageManager for CPU .. figure:: output_bounding-box_d6b70e_18_1.svg .. raw:: html
.. raw:: html
.. raw:: latex \diilbookstyleinputcell .. code:: python d2l.set_figsize() img = d2l.plt.imread('../img/catdog.jpg') d2l.plt.imshow(img); .. figure:: output_bounding-box_d6b70e_21_0.svg .. raw:: html
.. raw:: html
Bounding Boxes -------------- In object detection, we usually use a *bounding box* to describe the spatial location of an object. The bounding box is rectangular, which is determined by the :math:`x` and :math:`y` coordinates of the upper-left corner of the rectangle and the such coordinates of the lower-right corner. Another commonly used bounding box representation is the :math:`(x, y)`-axis coordinates of the bounding box center, and the width and height of the box. Here we define functions to convert between these two representations: ``box_corner_to_center`` converts from the two-corner representation to the center-width-height presentation, and ``box_center_to_corner`` vice versa. The input argument ``boxes`` should be a two-dimensional tensor of shape (:math:`n`, 4), where :math:`n` is the number of bounding boxes. .. raw:: html
pytorchmxnettensorflow
.. raw:: html
.. raw:: latex \diilbookstyleinputcell .. code:: python #@save def box_corner_to_center(boxes): """Convert from (upper-left, lower-right) to (center, width, height).""" x1, y1, x2, y2 = boxes[:, 0], boxes[:, 1], boxes[:, 2], boxes[:, 3] cx = (x1 + x2) / 2 cy = (y1 + y2) / 2 w = x2 - x1 h = y2 - y1 boxes = torch.stack((cx, cy, w, h), axis=-1) return boxes #@save def box_center_to_corner(boxes): """Convert from (center, width, height) to (upper-left, lower-right).""" cx, cy, w, h = boxes[:, 0], boxes[:, 1], boxes[:, 2], boxes[:, 3] x1 = cx - 0.5 * w y1 = cy - 0.5 * h x2 = cx + 0.5 * w y2 = cy + 0.5 * h boxes = torch.stack((x1, y1, x2, y2), axis=-1) return boxes .. raw:: html
.. raw:: html
.. raw:: latex \diilbookstyleinputcell .. code:: python #@save def box_corner_to_center(boxes): """Convert from (upper-left, lower-right) to (center, width, height).""" x1, y1, x2, y2 = boxes[:, 0], boxes[:, 1], boxes[:, 2], boxes[:, 3] cx = (x1 + x2) / 2 cy = (y1 + y2) / 2 w = x2 - x1 h = y2 - y1 boxes = np.stack((cx, cy, w, h), axis=-1) return boxes #@save def box_center_to_corner(boxes): """Convert from (center, width, height) to (upper-left, lower-right).""" cx, cy, w, h = boxes[:, 0], boxes[:, 1], boxes[:, 2], boxes[:, 3] x1 = cx - 0.5 * w y1 = cy - 0.5 * h x2 = cx + 0.5 * w y2 = cy + 0.5 * h boxes = np.stack((x1, y1, x2, y2), axis=-1) return boxes .. raw:: html
.. raw:: html
.. raw:: latex \diilbookstyleinputcell .. code:: python #@save def box_corner_to_center(boxes): """Convert from (upper-left, lower-right) to (center, width, height).""" x1, y1, x2, y2 = boxes[:, 0], boxes[:, 1], boxes[:, 2], boxes[:, 3] cx = (x1 + x2) / 2 cy = (y1 + y2) / 2 w = x2 - x1 h = y2 - y1 boxes = tf.stack((cx, cy, w, h), axis=-1) return boxes #@save def box_center_to_corner(boxes): """Convert from (center, width, height) to (upper-left, lower-right).""" cx, cy, w, h = boxes[:, 0], boxes[:, 1], boxes[:, 2], boxes[:, 3] x1 = cx - 0.5 * w y1 = cy - 0.5 * h x2 = cx + 0.5 * w y2 = cy + 0.5 * h boxes = tf.stack((x1, y1, x2, y2), axis=-1) return boxes .. raw:: html
.. raw:: html
We will define the bounding boxes of the dog and the cat in the image based on the coordinate information. The origin of the coordinates in the image is the upper-left corner of the image, and to the right and down are the positive directions of the :math:`x` and :math:`y` axes, respectively. .. raw:: html
pytorchmxnettensorflow
.. raw:: html
.. raw:: latex \diilbookstyleinputcell .. code:: python # Here `bbox` is the abbreviation for bounding box dog_bbox, cat_bbox = [60.0, 45.0, 378.0, 516.0], [400.0, 112.0, 655.0, 493.0] .. raw:: html
.. raw:: html
.. raw:: latex \diilbookstyleinputcell .. code:: python # Here `bbox` is the abbreviation for bounding box dog_bbox, cat_bbox = [60.0, 45.0, 378.0, 516.0], [400.0, 112.0, 655.0, 493.0] .. raw:: html
.. raw:: html
.. raw:: latex \diilbookstyleinputcell .. code:: python # Here `bbox` is the abbreviation for bounding box dog_bbox, cat_bbox = [60.0, 45.0, 378.0, 516.0], [400.0, 112.0, 655.0, 493.0] .. raw:: html
.. raw:: html
We can verify the correctness of the two bounding box conversion functions by converting twice. .. raw:: html
pytorchmxnettensorflow
.. raw:: html
.. raw:: latex \diilbookstyleinputcell .. code:: python boxes = torch.tensor((dog_bbox, cat_bbox)) box_center_to_corner(box_corner_to_center(boxes)) == boxes .. raw:: latex \diilbookstyleoutputcell .. parsed-literal:: :class: output tensor([[True, True, True, True], [True, True, True, True]]) .. raw:: html
.. raw:: html
.. raw:: latex \diilbookstyleinputcell .. code:: python boxes = np.array((dog_bbox, cat_bbox)) box_center_to_corner(box_corner_to_center(boxes)) == boxes .. raw:: latex \diilbookstyleoutputcell .. parsed-literal:: :class: output array([[ True, True, True, True], [ True, True, True, True]]) .. raw:: html
.. raw:: html
.. raw:: latex \diilbookstyleinputcell .. code:: python boxes = tf.constant((dog_bbox, cat_bbox)) box_center_to_corner(box_corner_to_center(boxes)) == boxes .. raw:: latex \diilbookstyleoutputcell .. parsed-literal:: :class: output .. raw:: html
.. raw:: html
Let’s draw the bounding boxes in the image to check if they are accurate. Before drawing, we will define a helper function ``bbox_to_rect``. It represents the bounding box in the bounding box format of the ``matplotlib`` package. .. raw:: html
pytorchmxnettensorflow
.. raw:: html
.. raw:: latex \diilbookstyleinputcell .. code:: python #@save def bbox_to_rect(bbox, color): """Convert bounding box to matplotlib format.""" # Convert the bounding box (upper-left x, upper-left y, lower-right x, # lower-right y) format to the matplotlib format: ((upper-left x, # upper-left y), width, height) 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) .. raw:: html
.. raw:: html
.. raw:: latex \diilbookstyleinputcell .. code:: python #@save def bbox_to_rect(bbox, color): """Convert bounding box to matplotlib format.""" # Convert the bounding box (upper-left x, upper-left y, lower-right x, # lower-right y) format to the matplotlib format: ((upper-left x, # upper-left y), width, height) 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) .. raw:: html
.. raw:: html
.. raw:: latex \diilbookstyleinputcell .. code:: python #@save def bbox_to_rect(bbox, color): """Convert bounding box to matplotlib format.""" # Convert the bounding box (upper-left x, upper-left y, lower-right x, # lower-right y) format to the matplotlib format: ((upper-left x, # upper-left y), width, height) 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) .. raw:: html
.. raw:: html
After adding the bounding boxes on the image, we can see that the main outline of the two objects are basically inside the two boxes. .. raw:: html
pytorchmxnettensorflow
.. raw:: html
.. raw:: latex \diilbookstyleinputcell .. code:: 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')); .. figure:: output_bounding-box_d6b70e_75_0.svg .. raw:: html
.. raw:: html
.. raw:: latex \diilbookstyleinputcell .. code:: 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')); .. figure:: output_bounding-box_d6b70e_78_0.svg .. raw:: html
.. raw:: html
.. raw:: latex \diilbookstyleinputcell .. code:: 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')); .. figure:: output_bounding-box_d6b70e_81_0.svg .. raw:: html
.. raw:: html
Summary ------- - Object detection not only recognizes all the objects of interest in the image, but also their positions. The position is generally represented by a rectangular bounding box. - We can convert between two commonly used bounding box representations. Exercises --------- 1. Find another image and try to label a bounding box that contains the object. Compare labeling bounding boxes and categories: which usually takes longer? 2. Why is the innermost dimension of the input argument ``boxes`` of ``box_corner_to_center`` and ``box_center_to_corner`` always 4? .. raw:: html
pytorchmxnet
.. raw:: html
`Discussions `__ .. raw:: html
.. raw:: html
`Discussions `__ .. raw:: html
.. raw:: html