ImageAnalyst

ImageAnalyst is a library that simplifies image analysis. The library achieves this goal by standardizing the input and output vectors of a few machine learning models and by providing some high-level analysis algorithms.

Requirements

The application requires:

• Python ~= 3.9
• pip

Extras

The application has some extras that can be installed:

• cv2
• hf
• tf
• onnx

Download & Installation

There is two ways to download and install the application.

Using PyPI

You can download and install the application using PyPI. To do so, run the following command:

pip install image-analyst

Using the GitHub releases

You can download the application on the downloads page. Then, you can install the application by running the following command:

pip install image_analyst-X.X.X-py3-none-any.whl

(Note: The X.X.X must be replaced by the version that you want to install.)

Example

This example allows you to track objects from your webcam. It requires the cv2 extra.

from image_analyst.cv2.utils import convert_image, create_frame_generator
from image_analyst.cv2.models import YoloV3OpenCV
from image_analyst.trackers import IOUTracker
import cv2

def report_callback(filename: str, current_size: float, total_size: float):
    print("{} {:.2f}%".format(filename, current_size/total_size*100), end="\r", flush=True)

model = YoloV3OpenCV.coco(report_callback=report_callback)

tracking_function = IOUTracker(model)

with create_frame_generator(0) as frame_generator:
    for frame in frame_generator:
        converted_frame = convert_image(frame, model.supported_format, model.supported_dtype)
        instances = tracking_function(converted_frame)

        for instance in instances:
            xmin, ymin, xmax, ymax = instance.bounding_box.as_tuple()

            text = "{} {} {:.2f}".format(instance.id, instance.class_name, instance.score)
            cv2.rectangle(frame, (xmin, ymin), (xmax, ymax), (0, 255, 0), 2)
            cv2.putText(frame, text, (xmin, ymin), cv2.FONT_HERSHEY_SIMPLEX, 1, (0, 255, 0), 2)

        cv2.imshow("Tracking", frame)

        if cv2.waitKey(1) & 0xFF == ord('q'):
            break

cv2.destroyAllWindows()

License

All code is licensed for others under a MIT license (see LICENSE).