Welcome to my object detection project

This project is intended to be a fun project for learning about object detection.

REFERENCE

Thanks to these resources:

(tutorial): https://www.youtube.com/watch?v=WgPbbWmnXJ8&t=2234s
(tracker algorithm): https://github.com/abewley/sort

Use case: step by step

1. Clone this project to your local machine.

Note : the location of this repository will be downloaded on the folder that you are currently on, so you should put it in the place where you can run import the package easily.

git clone https://github.com/OuiSinthala/LaoODetection.git

2. Create a new virtual environment to avoid package conflict

macos or linux:

python3 -m venv __your virtual environment name__

windows:

python -m venv __your virtual environment name__

3. Activate the virtual environment

the " " are not included

macos or linux:

source "your virtual environment"/bin/activate

windows:

"your virtual environment"/Scripts/activate

4. Install necessary dependencies for this package to the virtual environment.

Once you have created the virtual environment, you should run this command.

pip install -r requirements.txt

5. Finally: If there is no any error, that is mean that you have successfully installed the package.

Next are some examples:

1. People detection

In this example, I only test with detecting only people.
You can try other objects if there is any error:
Tell me :) and fix it together.

import cv2 as cv
from LaoODetection import ObjectDetector    #import the package

def main():
    # Input source: webcam
    video_capture = cv.VideoCapture(0)

    # chose object target
    detector = ObjectDetector(target_object="person") 
    # run the object detection algorithm
    detector.object_detection(video_capture)    
    # getting data from the algorithm
    print("Detected: {0} people.".format(detector.number_of_detected_objects))

if __name__ == "__main__":
    main()

2. Counting specific object on a specific area:

In this example we need the region of interest in order to detect only specific location in the frame
The target_object can be other objects such as person, cat, dog

import cv2 as cv
from LaoODetection import ObjectDetector

COUNTER_line1 = [80, 400, 700, 400]  # this should be changed to fit your need
region_of_interest = cv.imread("RegionOfInterest/roi_people.png")


def main():
    # Note you can change the input source to be a webcam
    # Input source: Video
    video_location = 'testing_video/cars.mp4'
    video_capture = cv.VideoCapture(video_location)

    # Running the object detection algorithm
    target = ["car", "bus", "truck", "motorbike"]
    detector = ObjectDetector(target_object=target)
    detector.object_line_counter(video_capture=video_capture, region_of_interest=region_of_interest,
                                 counter_line1=COUNTER_line1, detecting_range=(15, 30))
    # getting the result to command screen
    print("Detected: {0} people.".format(detector.number_of_detected_objects))


if __name__ == "__main__":
    main()

ouisinthala / laoodetection Goto Github PK

laoodetection's Introduction

Welcome to my object detection project

This project is intended to be a fun project for learning about object detection.

REFERENCE

Thanks to these resources:

Use case: step by step

1. Clone this project to your local machine.

Note : the location of this repository will be downloaded on the folder that you are currently on, so you should put it in the place where you can run import the package easily.

2. Create a new virtual environment to avoid package conflict

macos or linux:

windows:

3. Activate the virtual environment

the " " are not included

macos or linux:

windows:

4. Install necessary dependencies for this package to the virtual environment.

Once you have created the virtual environment, you should run this command.

5. Finally: If there is no any error, that is mean that you have successfully installed the package.

Next are some examples:

1. People detection

In this example, I only test with detecting only people.

You can try other objects if there is any error:

Tell me :) and fix it together.

2. Counting specific object on a specific area:

laoodetection's People

Contributors

Watchers

Recommend Projects

Recommend Topics

Recommend Org