Skip to content

A set of libraries for encapsulating smart vision algorithms

License

Notifications You must be signed in to change notification settings

opencv/open_vision_capsules

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

OpenVisionCapsules

Documentation Status

This repository contains the OpenVisionCapsules SDK, a set of Python libraries for encapsulating machine learning and computer vision algorithms for intelligent video analytics.

Encapsulating an algorithm allows it to be deployed as a single, self-describing file that inputs and outputs data in a standard format. This makes deployment and integration significantly easier than starting with a model file or a snippet of source code. Capsules are descriptive of their input and output requirements, allowing OpenVisionCapsules to route data between capsules automatically.

This project is split into two packages, vcap and vcap-utils. vcap contains the necessary facilities to create and encapsulate an algorithm. vcap-utils contains a set of utilities that make encapsulating algorithms of certain types easier.

Project Status

OpenVisionCapsules is in a developer preview phase. We're looking for developer feedback before reaching a stable 1.0 release. If you find any bugs or have suggestions, please open an issue.

Getting Started

Take a look at the documentation here.

A couple example capsules are available under vcap/examples, demonstrating how to create classifier and detector capsules from TensorFlow models.

Installation

To install OpenVisionCapsules locally, clone the repository and run the following commands to install the vcap and vcap-utils packages in the current environment.

pip3 install -e ./vcap
pip3 install -e ./vcap_utils

Examples

To make use of the example capsules in the vcap/examples/ directory, make sure to run the tests with pytest (from the root of the repo). The tests download all the necessary models and images, including the models for the example capsules.

Make sure vcap & vcap-utils installation is done before tests,

pytest -v -x .

A repository of open source capsules can be found here.

capsule inference test,

python3 tools/openvisioncapsule_tools/capsule_infer/capsule_infer.py --capsule ../capsule-zoo/capsules/detector_person_vehicle_bike_openvino  --images tests/test_images/two_people.jpg