tomsriddle/cv-objectdetection repository overview

⁠Graphical Degrading and Object Detection Assignment

⁠Overview

The Git repository (https://github.com/shaileshhemdev/hemdev-705.603Spring24/tree/master/cvbasicsAssignment-main⁠) contains an implementation of the OpenCV2 Deep Neural Network using YOLO model for object detection in images. It has the following key files

• Notebook (ObjectDetection.ipynb): In this notebook we demonstrate how certain reduction techniques on images can impact the object detection
• Python Class (object_detection.py): This class encapsulates the Deep Neural Network implementation to return the objects detected on a given image with the corresponding confidences and bounding boxes
• Flask Service (object_detection_service.py): This provides a REST service to post any image to the service to obtain the objects detected along with their confidence

⁠Local Development Environment used for building image

• Apple Mac M1 chip
• Sonoma 14.1.2
• Docker Desktop for Mac 4.26.1 (131620)

⁠Running on Local

• You need git, python 3.8 and pip. See https://pip.pypa.io/en/stable/installation/⁠

• It is recommended that you use Python virtual environments. See https://www.freecodecamp.org/news/how-to-setup-virtual-environments-in-python/⁠

• Run pip3 install -r requirements.txt

• Run python object_detection_service.py

⁠How to use the image using Docker

⁠Pull Image

To pull an image use the following

docker pull tomsriddle/cv-objectdetection:1.0

After pulling the image check that it is present using following

docker image ls

⁠Build Image from Local

docker buildx build -t "tomsriddle/cv-objectdetection:1.0" --load --platform linux/amd64,linux/arm64 .

⁠Run Image

To run the image use following

docker run -p <host port>:8786 -v <host path>:/workspace/shared-data -e data-folder=/workspace/shared-data/ "tomsriddle/cv-objectdetection:1.0" 

Note: See the volume mapping - this is needed for the saving of the image for the POST API. This volume should have the yolov3.cfg and yolov3.weights files as shown below

⁠Docker Image and Run Example

⁠API Usage

Following APIs exposed

⁠Detect

http://localhost:8786/detect

We need to submit form-data with image submitted as imagefile attribute

⁠Detect Example

⁠Detect Example in Docker

Notice how when you make the call the sent image is saved on your volume