Greetings, fellow developers! 👋
Empower your creativity with Mech-DLK SDK, ready for you to unleash your programming prowess. Start crafting your own applications today!
Mech-DLK SDK is a secondary development software kit specifically designed to be used with Mech-DLK. It mainly helps you easily do deep learning inference in your software systems. With Mech-DLK SDK, you can rapidly deploy deep learning models and flexibly integrate deep learning functionality into your own applications without reliance on Mech-Vision. Currently, development in C#, C++, and C languages is supported.
📥 If you need to install Mech-DLK, please go to Download Center to get the Mech-DLK Installer. You can contact us at info@mech-mind.net to learn more about software licenses.
NOTE: If you are using Mech-DLK SDK version 2.1.0 or later, please download Mech-DLK version 2.6.0 or later. If you are using Mech-DLK SDK version earlier than 2.1.0, please download Mech-DLK version 2.4.2 - 2.5.4.
💡 If you have any questions or have anything to share regarding our SDK, feel free to post on Mech-Mind Online Community.
- Make sure that you've purchased Mech-DLK's Pro-Train/Pro-Run software license.
- Make sure that you've downloaded and installed the Sentinel LDK encryption driver.
- Make sure that you've activated or updated the software license that you purchased.
- It is recommended that the used device should satisfy the following requirements:
| Authorized software license version | Pro-Run | Pro-Train |
|---|---|---|
| Operating system | Windows 10 or above | Windows 10 or above |
| CPU | Intel® Core™ i7-6700 or above | Intel® Core™ i7-6700 or above |
| Memory | 8GB or above | 16GB or above |
| Graphics card | GeForce GTX 1660 or above | GeForce RTX 3060 or above |
| Graphics card driver | Version 472.50 or above | Version 472.50 or above |
-
Create a local project folder on your device, such as dlk_sdk.
-
Clone the repository of Mech-DLK SDK to the project folder.
-
Download the third-party libraries (3rd_dll.zip) and resources (resources.zip) that Mech-DLK SDK relies on to the project folder.
-
Unzip the downloaded packages of third-party libraries and resources.
- CRC32 value for third-party libraries: 9037EC58
- CRC32 value for resources: 3C23BC3A
Two categories of samples are provided: Basic and Advanced.
Samples using models exported from Mech-DLK to do inference of single images and simultaneous inference of images as well as obtain and visualize results.
-
ImageInfer: a sample for inference of single images (both single models and cascaded models are supported).
-
MultiImageInfer: a sample for simultaneous inference of images (both single models and cascaded models are supported).
Samples demonstrating collaborative development of Mech-DLK SDK with HALCON/OpenCV.
-
ImageInferWithHalcon: a sample that runs on the basis of Mech-DLK SDK and HALCON.
-
ImageInferWithOpenCV: a sample that runs on the basis of Mech-DLK SDK and OpenCV.
Two categories of samples are provided: Basic and Advanced.
Samples using single models exported from Mech-DLK to do inference of single images and simultaneous inference of multiple images as well as obtain and visualize results.
-
ImageInfer: a sample for inference of single images.
-
MultiImageInfer: a sample for simultaneous inference of images.
A sample demonstrating collaborative development of Mech-DLK SDK with OpenCV.
- ImageInferWithOpenCV: a sample running on the basis of Mech-DLK SDK and OpenCV.
Two categories of samples are provided: Basic and Advanced.
Samples using single models exported from Mech-DLK to do inference of single images as well as obtain and visualize results.
-
Classification: a sample for inference based on the Classification model.
-
DefectSegmentation: a sample for inference based on the Defect Segmentation model.
-
FastPositioning: a sample for inference based on the Fast Positioning model.
-
InstanceSegmentation: a sample for inference based on the Instance Segmentation model.
-
ObjectDetection: a sample for inference based on the Object Detection model.
-
TextDetection: a sample for inference based on the Text Detection model.
-
TextRecognition: a sample for inference based on the Text Recognition model.
-
UnsupSegmentation: a sample for inference based on the Unsupervised Segmentation model.
Samples for simultaneous inference of multiple images and inference based on cascaded models.
-
CascadeModel: a sample for inference based on cascaded models.
-
FolderImagesInfer: a sample used to show the inference of images in a folder one by one.
-
MultiImageInfer: a sample for simultaneous inference of images.
Please refer to the Get Started section for instructions on using Mech-DLK SDK for model inference.
You may also find other contents that can help you get started with Mech-DLK SDK.
- System requirements
- How to build and run samples
- Mech-DLK SDK C# APIs
- Mech-DLK SDK C++ APIs
- Mech-DLK SDK C APIs
- FAQs
The above samples of Mech-DLK SDK are distributed under the BSD license.