人工智能ImageAI的一系列示范代码，教你怎么使用ImageAI

# ImageAI <br> A python library built to empower developers to build applications and systems with self-contained Deep Learning and Computer Vision capabilities using simple and few lines of code. <br><br> <img src="logo1.png" style="width: 500px; height: auto; margin-left: 50px; " /> <br> An <b>AI Commons</b> project <a href="https://commons.specpal.science" >https://commons.specpal.science </a> Developed and Maintained by [Moses Olafenwa](https://twitter.com/OlafenwaMoses) and [John Olafenwa](https://twitter.com/johnolafenwa), brothers, creators of [TorchFusion](https://github.com/johnolafenwa/TorchFusion) and Authors of [Introduction to Deep Computer Vision](https://john.specpal.science/deepvision) <hr> Built with simplicity in mind, <b>ImageAI</b> supports a list of state-of-the-art Machine Learning algorithms for image prediction, custom image prediction, object detection, video detection, video object tracking and image predictions trainings. <b>ImageAI</b> currently supports image prediction and training using 4 different Machine Learning algorithms trained on the ImageNet-1000 dataset. <b>ImageAI</b> also supports object detection, video detection and object tracking using RetinaNet, YOLOv3 and TinyYOLOv3 trained on COCO dataset. <br> Eventually, <b>ImageAI</b> will provide support for a wider and more specialized aspects of Computer Vision including and not limited to image recognition in special environments and special fields. <br> <br> <b>New Release : ImageAI 2.0.2</b> <br> What's new: <br> - Option to state image size during custom image prediction model trainings <br> - Object Detection and Video Object detection now returns bounding box coordinates **('box points')** (x1,y1,x2, y2) for each object detected in addition to object's 'name' and 'percentage probability' <br> - Options to hide 'percentage probability' and/or object 'name' from being shown in detected image or video - Support for video object detection on video live stream from device camera, connected camera and IP camera <br> - Support for **YOLOv3** and **TinyYOLOv3** for all object detection and video object detection tasks. - Video object detection for all input types (video file and camera) now allows defining custom functions to execute after each frame, each second and each minute of the video is detected and processed. Also include option to specify custom function at once video is fully detected and processed <br> - For each custom function specified, **ImageAI** returns the **frame**/**seconds**/**minute**/**full video analysis** of the detections that include the objects' details ( **name** , **percentage** **probability**, **box_points**), number of instance of each unique object detected (counts) and overall average count of the number of instance of each unique object detected in the case of **second** / **minute** / **full video analysis**<br> - Options to return detected frame at every frame, second or minute processed as a **Numpy array**. <br> <br> <br> <h3><b><u>TABLE OF CONTENTS</u></b></h3> <a href="#dependencies" >&#9635 Dependencies</a><br> <a href="#installation" >&#9635 Installation</a><br> <a href="#prediction" >&#9635 Image Prediction</a><br> <a href="#detection" >&#9635 Object Detection</a><br> <a href="#videodetection" >&#9635 Video Object Detection, Tracking & Analysis</a><br> <a href="#customtraining" >&#9635 Custom Model Training</a><br> <a href="#customprediction" >&#9635 Custom Image Prediction</a><br> <a href="#documentation" >&#9635 Documentation</a><br> <a href="#sample" >&#9635 Projects Built on ImageAI</a><br> <a href="#recommendation" >&#9635 AI Practice Recommendations</a><br> <a href="#support" >&#9635 Support the ImageAI Project</a><br> <a href="#contact" >&#9635 Contact Developers</a><br> <a href="#contributors" >&#9635 Contributors</a><br> <a href="#ref" >&#9635 References</a><br> <br><br> <div id="dependencies"></div> <h3><b><u>Dependencies</u></b></h3>To use <b>ImageAI</b> in your application developments, you must have installed the following dependencies before you install <b>ImageAI</b> : <br> <br> <span><b>- Python 3.5.1 (and later versions) </b> <a href="https://www.python.org/downloads/" style="text-decoration: none;" >Download</a> (Support for Python 2.7 coming soon) </span> <br> <span><b>- pip3 </b> <a href="https://pypi.python.org/pypi/pip" style="text-decoration: none;" >Install</a></span> <br> <span><b>- Tensorflow 1.4.0 (and later versions) </b> <a href="https://www.tensorflow.org/install/install_windows" style="text-decoration: none;" > Install</a></span> or install via pip <pre> pip3 install --upgrade tensorflow </pre> <span><b>- Numpy 1.13.1 (and later versions) </b> <a href="https://www.scipy.org/install.html" style="text-decoration: none;" >Install</a></span> or install via pip <pre> pip3 install numpy </pre> <span><b>- SciPy 0.19.1 (and later versions) </b> <a href="https://www.scipy.org/install.html" style="text-decoration: none;" >Install</a></span> or install via pip <pre> pip3 install scipy </pre> <span><b>- OpenCV </b> <a href="https://pypi.python.org/pypi/opencv-python" style="text-decoration: none;" >Install</a></span> or install via pip <pre> pip3 install opencv-python </pre> <span><b>- Pillow </b> <a href="https://pypi.org/project/Pillow/2.2.1/" style="text-decoration: none;" >Install</a></span> or install via pip <pre> pip3 install pillow </pre> <span><b>- Matplotlib </b> <a href="https://matplotlib.org/users/installing.html" style="text-decoration: none;" >Install</a></span> or install via pip <pre> pip3 install matplotlib </pre> <span><b>- h5py </b> <a href="http://docs.h5py.org/en/latest/build.html" style="text-decoration: none;" >Install</a></span> or install via pip <pre> pip3 install h5py </pre> <span><b>- Keras 2.x </b> <a href="https://keras.io/#installation" style="text-decoration: none;" >Install</a></span> or install via pip <pre> pip3 install keras </pre> <div id="installation"></div> <h3><b><u>Installation</u></b></h3> To install ImageAI, run the python installation instruction below in the command line: <br><br> <span> <b>pip3 install https://github.com/OlafenwaMoses/ImageAI/releases/download/2.0.2/imageai-2.0.2-py3-none-any.whl </b></span> <br><br> <br> or download the Python Wheel <a href="https://github.com/OlafenwaMoses/ImageAI/releases/download/2.0.2/imageai-2.0.2-py3-none-any.whl" ><b> imageai-2.0.2-py3-none-any.whl</b></a> and run the python installation instruction in the command line to the path of the file like the one below: <br><br> <span> <b>pip3 install C:\User\MyUser\Downloads\imageai-2.0.2-py3-none-any.whl</b></span> <br><br> <div id="prediction"></div> <h3><b><u>Image Prediction</u></b></h3> <p><img src="images/1.jpg" style="width: 400px; height: auto;" /> <pre>convertible : 52.459555864334106 sports_car : 37.61284649372101 pickup : 3.1751200556755066 car_wheel : 1.817505806684494 minivan : 1.7487050965428352</pre> </p> <b>ImageAI</b> provides 4 different algorithms and model types to perform image prediction, trained on the ImageNet-1000 dataset. The 4 algorithms provided for image prediction include <b>SqueezeNet</b>, <b>ResNet</b>, <b>InceptionV3</b> and <b>DenseNet</b>. Click the link below to see the full sample codes, explanations and best practices guide. <a href="imageai/Prediction/" ><button style="font-size: 20px; color: white; background-color: steelblue; height: 50px; border-radius: 10px; " > >>> Tutorial & Guide </button></a> <br> <br> <div id="detection"></div> <h3><b><u>Object Detection</u></b></h3> <div style="width: 600px;" > <b><p><i>Input Image</i></p></b></br> <img src="images/image2.jpg" style="width: 500px; height: auto; margin-left: 50px; " /> <br>