Making Machine Vision Simple
Machine vision has been used widely in all factories and is a critical process to ensure that correct parts with good quality are shipped out to their customers. It is no more an option to have machine vision, but a definite choice to help to minimize the job load for most of the Quality Assurance engineer.
The benefits of automated inspection outweigh human labour in the aspects of consistency & increased output. There have been many areas where machine visions are used with our realization. Some examples will include code reading in the warehouse or production floor for traceability purpose, or to act as the “eyes” for the robots. Measurements or integrity check of the products, and defects inspection are commonly used in different stages of the production of the final product.
Machine vision technology has been progressing & changing significantly over the years. Many new solutions, including line scan, 3D, thermal imaging have been introduced to solve some of the challenging problems in the industry. Over the recent years, AI deep learning has also been used widely for defects inspection, sorting & OCR applications too. Nonetheless, certain fundamentals must be understood in order to achieve successful implementation of any machine vision application.
As we break down the different components of a vision solution, it will be classified as camera, illumination, lens & software. Each of the components are important to ensure the success and repeatability on the outcome of any projects.
It will be useful if we adopt a different approach in looking at machine vision. We can assess and classify projects that are simple or complex.
- A simple project can be straight-forward applications which includes code reading, OCR, alignment or measurement. The project does not need much evaluation or required customized lens or illumination & can be using up to 2 cameras. For such projects, using an integrated vision controller with built-in ready-to-use application software will be a better & cost-effective approach to guide the user on the steps to fulfil a requirement quickly without much hustle.
- A complex project will be those requirements that involves intensive engineering time & selecting customized hardware to fulfil the requirement. The complexity may be due to the surface finishing of the inspected parts where it is difficult to achieve the desired “ideal” image to show a good contrast. It can also be the inspection requirement where standard imaging method might not be able to fulfil the goals. While It may require special camera, illumination or lens to achieve a good image, this will involve good experience & intensive engineering evaluation of many product samples. For such projects, there may be a need to use specialized tools within the vision software to solve the problem.
To meet the needs of the system integrators and users, Wizer vision software comes in different configuration & modules.
- This is the basic platform which form the basis of all our other modules. There is no programming required and is a drag & drop of different tools to create the solution.
- With proper training & experience, many simple & complex problems can be resolved with this software
Wizer RTU (Ready-to-Use)
- As there are many vision applications in the factory that are straight forward & simple, there is no need for the end users to carry out customization. Wizer software has the option of the RTU (Ready-to-Use modules) for the users to do their setup with minimal knowledge or training. The GUI (Graphic User Interface) has been designed in simplicity to speed up the set-up in the shortest time.
- This module comprised of more than 20 different applications, e.g. code reader, OCR, measurement, alignment, etc Depending on the project requirement, the user selects the required module and a ready-to-use (RTU) comprehensive GUI will pop up for the user to start the project
- The user will follow the steps to complete the whole process of capturing the image, filtering, alignment, inspection & communication
As standard of living improved in 1st & 2nd tier cities, cost of labour has increased too. Retaining or employing operators or engineers in factories are getting more difficult as there are more choices for the employees. To increase production output, improve quality level & to solve some of the employment issues, where possible, the manufacturing industry are moving towards fully automated operations with minimal manpower. At the same time, we have to simplify the technological implementation for users to have the ease of use.
Wizer vision software has been designed with different modules to handle simple & complex requirement, with cost benefits for the users. The users will have better control of the technology during the implementation & changes between the product models in the production.
You can view our website www.vizcam.net or to write to us [email protected] to understand how we can help you to have a different approach towards solving inspection problem.
Wizer DP (AI Deep Learning)
- The AI deep learning is integrated within Wizer for simple setup & operation.
- The training of the images will be done externally, and the trained recipe will be called into the Wizer DP easily without any difficulty or complex steps.
Wizer Intuit (AI Machine Learning)
- Wizer Intuit is our AI machine learning software where the user will train good images and take short training time (e.g. less than 10 sec for 50 good images). The AI will learn the characteristic of the good parts, and the user can then set conditions for the inspection requirement.
- This will be suitable for inspection requirement where there is minimal lens distortion in the captured image & minimal variation between parts.
- For a printed brochure inspection, the whole process to build up the solution can take 20 to 60 mins.
- Wizer RTU has a module for Wizer Intuit to simplify the steps to build up the solution.