PyDataProcessΒΆ
Module offering implementation of specialization classes to handle inputs/outputs and tasks involved in Ikomia workflows for concrete use cases. It consists of Python bindings from C++ Ikomia Core.
In addiction to workflow management, the system uses factory design pattern to allow integration of third-party plugins. To this end, when you want to add your own plugin, you have to implement (override) three factory classes derived from the following:
CPluginProcessInterface
: abstract base class exposing the two factories required by the plugin engine (task and widget)
CTaskFactory
: abstract base class for process instanciation
CWidgetFactory
: abstract base class for widget instanciation
This module provides class specialization for several types of usual inputs/outputs. It also provides class specialization for common processing task. You will find below details about implementation of such specializations.
Plugins connection
Abstract class defining the core structure of a plugin interface. |
|
Abstract class defining the core structure of the task factory. |
|
Abstract class defining the core structure of the widget factory. |
Input/Output management
Define multi-dimensional array as input or output of a task. |
|
Define input or output for a task consuming or producing measures on blobs (ie connected components). |
|
Virtual base classe to define task input or output containing deep learning dataset structure. |
|
Define input or output for a task consuming or producing data as numeric values. |
|
Define task input containing graphics items. |
|
Define task output containing graphics items. |
|
Define task input or output for a task dedicated to image processing (2D or 3D). |
|
Store single instance segmentation information (class properties): class index, label, confidence, box, mask and color. |
|
Define input or output managing common information extracted by instance segmentation task. |
|
Store single object keypoints information (class properties): label, confidence, box, color and points. |
|
Store link information between two keypoints (class properties): starting point index, ending point index, label and color. |
|
Define input or output managing common information extracted by keypoints detection task. |
|
Enum - List of a display types for numeric values |
|
Define input or output for a task consuming or producing data as numeric values. |
|
Store values of a given measure computed on a single blob of an image. |
|
Store single object detection information (class properties): label, confidence, box and color. |
|
Define input or output managing common information extracted by object detection task. |
|
Define input or output for a task to handle file or folder path. |
|
Define input or output managing common information extracted by semantic segmentation task. |
|
Store single text field information (class properties): label, text, confidence, polygon and color. |
|
Define input or output managing common information extracted by OCR task. |
|
Define an input or output for a task dedicated to video management. |
|
Define a widget output for a task. |
|
Enum - List of plot types |
Task management
Implement all basic features used in a task dedicated to 2D image processing. |
|
Add user interactions capability to a 2D image process task. |
|
Base class for classification task in Computer Vision. |
|
Internal use only. |
|
Algorithms registry of the Ikomia platform. |
|
Base class for instance segmentation task in Computer Vision. |
|
Base class for keypoint detection task in Computer Vision. |
|
Base class for object detection task in Computer Vision. |
|
Base class for semantic segmentatio task in Computer Vision. |
|
Manage metadata associated with a task. |
|
Add video specific features. |
|
Add optical flow specific features for methods based on OpenCV framework. |
|
Add specific features for tracking task. |
|
Workflow management of Computer Vision tasks. |
Widget management
Abstract class defining the core structure of the widget factory. |