Overview of Mech-DLK

Brief Description

Mech-DLK is a deep learning platform software independently developed by Mech-Mind Robotics. With a variety of built-in industry-leading AI algorithms and through intuitive and simple UI interactions, it helps customers solve complex problems, including overlapping object recognition, highly difficult defect detection, and product grading and classification, etc. It can improve production efficiency, product yield, and reduce production line labor costs, and is suitable for industries including consumer electronics, new energy, automobiles, home appliances, logistics, etc.

Introduction of the Module Feature

Classification: Classify multiple types of objects. Given a limited amount of images of different types of objects, this module can classify different types of objects by deep learning.

Object Detection: Quickly detect objects in images. Given a limited amount of images, this module can detect and locate target objects.

Semantic Segmentation: Detect the defects in images. Given a certain amount of positive (OK) samples and negative (NG) samples, this module can detect the defects in images via deep learning.

Instance Segmentation: Segment and locate the target objects in images and classify them. Given a limited amount of images, this module can locate different objects and classify them via deep learning.

Attention

All of the above modules can be used separately to train deep learning models that meet your needs. If you have customized requirements for a variety of functions, Mech-DLK can also meet your needs by cascading different modules for training.

Custom Development Configuration Requirements

Operating system	Windows 7 and above
Development platform	VS2013, VS2015, VS2017, VS2019 (Recommended)
Programming language	C, C++, C#