In another post, we explained how to apply Object Detection in Tensorflow. In this post, we will provide some examples of how you can apply Object Detection using the YOLO algorithm in Images and Videos. For our example, we will use the ImageAI Python library where with a few lines of code we can apply object detection.
Below we represent the code for Object Detection in Images.
from imageai.Detection import ObjectDetection
import osexecution_path = os.getcwd()
detector = ObjectDetection()
detector.setModelTypeAsYOLOv3()
detector.setModelPath( os.path.join(execution_path , "yolo.h5"))
detector.loadModel()detections = detector.detectObjectsFromImage(input_image=os.path.join(execution_path , "cycling001.jpg"), output_image_path=os.path.join(execution_path , "new_cycling001.jpg"), minimum_percentage_probability=30)for eachObject in detections: print(eachObject["name"] , " : ", eachObject["percentage_probability"], " : ", eachObject["box_points"] )print("--------------------------------")
And we get:
car : 99.66793060302734 : [395, 248, 701, 405]
--------------------------------
bicycle : 66.10226035118103 : [81, 270, 128, 324]
--------------------------------
bicycle : 99.86441731452942 : [242, 351, 481, 570]
--------------------------------
person : 99.92108345031738 : [269, 186, 424, 540]
--------------------------------
We also represent the Original and the Detected Image
Notice that it was able to detect the bicycle behind-A-M-A-Z-I-N-G!!!
Let’s provide another example of the Original and the Detected image
Notice that it detected the bed, the laptop and the two persons!
1. Microsoft Azure Machine Learning x Udacity — Lesson 4 Notes
2. Fundamentals of AI, ML and Deep Learning for Product Managers
3. Roadmap to Data Science
4. Work on Artificial Intelligence Projects
Assume that you have a video in your PC called “Traffic.mp4”, then by running this code you will be able to get the detected objects:
from imageai.Detection import VideoObjectDetection
import osexecution_path = os.getcwd()
detector = VideoObjectDetection()
detector.setModelTypeAsYOLOv3()
detector.setModelPath( os.path.join(execution_path , "yolo.h5"))detector.loadModel()video_path = detector.detectObjectsFromVideo(input_file_path=os.path.join(execution_path, "Traffic.mp4"),output_file_path=os.path.join(execution_path, "New_Traffic"), frames_per_second=20, log_progress=True)print(video_path)
And the detected video is here:
Let’s provide another example of a Video:
The following examples show how we can use our USB camera for Object Detection:
from imageai.Detection import VideoObjectDetectionimport os
import cv2execution_path = os.getcwd()
camera = cv2.VideoCapture(0)
detector = VideoObjectDetection()
detector.setModelTypeAsYOLOv3()
detector.setModelPath(os.path.join(execution_path , "yolo.h5"))
detector.loadModel()video_path = detector.detectObjectsFromVideo(camera_input=camera,output_file_path=os.path.join(execution_path, "camera_detected_video"), frames_per_second=20, log_progress=True, minimum_percentage_probability=30)print(video_path)
Below I represent just a snapshot of the recorded video of my office while I was coding. As you can see it was able to detect the books of the library behind me!
Credit: BecomingHuman By: Predictive Hacks
