Using this library you can detect objects from a video, images, with few lines of code
Project description
In this Project I am Using YoloV4 for object detection. To detect object using YoloV4 and OpenCv you have to write many lines of code, but using this framework you can do this using few lines of code.
Installation
pip install object_detector
Extract Images from a video
Import ExtractImages
class from object_detector
Library
>>> from object_detector import ExtractImages
Create a object ext
of ExtractImages
class
>>> ext = ExtractImages(path=0)
path=0
if you want to extract images from a video, write video path in place of 0.
or you can create object specifing Output path
>>> ext = ExtractImages(path=0,op=MyOutput)
By default Output path is Output
Start Extracting Images.
>>> ext.extract("A","B","C")
A, B, C are the path where extracting images collected. Relative path of "A" is MyOutput/A
Note: We are extracting Images in multiple path such that we can run our object detection task simultaneously.
Object Detection
Import ObjectDetector
class from object_detector
Library
>>> from object_detector import ObjectDetector
Create a object obj
of class ObjectDetector
>>> obj = ObjectDetector(weights="cars.weights", cfg="yolov4-custom.cfg", classes=['licence'])
Object detection is done using YoLo. cars.weights
is the trained file generated by training image dataset by darknet, and we are using yolov4-custom.cfg
.
>>> img = obj.detect_object(path)
Write path of the image, from which you want to detect object.
img
is numpy array of image. To the image from numpy array write
>>> import cv2
>>> cv2.imshow("Image_Name",img)
or You can write
>>> import matplotlib.pyplot as plt
>>> plt.imshow(img)
If you want image without label write the code as
>>> img = obj.detect_object(path,label=False)
If you want to img
only when obj is detected from image write the code as
>>> img = detect_object(path,detected_only=True)
To read text on the image add following code
>>> text = extract_text()
Show or Write objects from a video or camera
Import LiveDetect class and make a object of this class.
l = LiveDetect(weights="cars.weights", cfg="yolov4-custom.cfg", classes=['licence'])
Save images when object detect on video.
l.write_from_video(path="Traffic.mp4")
Show images when object detect on video.
l.show_from_video(path="Traffic.mp4")
Save images when object detect on the dir.
l.write_from_dir(path="Traffic.mp4")
Show images when object detect on the dir.
l.show_from_dir(path="Traffic.mp4")
Project details
Download files
Download the file for your platform. If you're not sure which to choose, learn more about installing packages.
Source Distribution
Built Distribution
File details
Details for the file object_detector-1.4.tar.gz
.
File metadata
- Download URL: object_detector-1.4.tar.gz
- Upload date:
- Size: 4.2 kB
- Tags: Source
- Uploaded using Trusted Publishing? No
- Uploaded via: twine/3.7.1 importlib_metadata/4.8.2 pkginfo/1.8.2 requests/2.26.0 requests-toolbelt/0.9.1 tqdm/4.62.3 CPython/3.6.15
File hashes
Algorithm | Hash digest | |
---|---|---|
SHA256 | c41994c9bb3977aa33be5114888658bce07fed711ff0bc82144feb6acb125835 |
|
MD5 | 2be41f8ed6d38eb2cc2964451eb76943 |
|
BLAKE2b-256 | 9a5bc5bc314e7c94e29397e467c4f5e997bace76a05a39f0b8113cbce7f609c7 |
Provenance
File details
Details for the file object_detector-1.4-py3-none-any.whl
.
File metadata
- Download URL: object_detector-1.4-py3-none-any.whl
- Upload date:
- Size: 4.5 kB
- Tags: Python 3
- Uploaded using Trusted Publishing? No
- Uploaded via: twine/3.7.1 importlib_metadata/4.8.2 pkginfo/1.8.2 requests/2.26.0 requests-toolbelt/0.9.1 tqdm/4.62.3 CPython/3.6.15
File hashes
Algorithm | Hash digest | |
---|---|---|
SHA256 | 5f439e097e9d4dcd1a817f1c33b99f752acad9df370505f599249861e632c111 |
|
MD5 | 25ebff94cbb32f31b58d965e5d8677d1 |
|
BLAKE2b-256 | f1e133fe21b3d37ef520dcdc5a5d6bfc3506d72b01f06c7cf076477f1944cceb |