Machine vision technology has progressed tremendously over the last ten years. The most typical use cases are automatic inspection, object detection, event identification, process control and guidance. We design and implement custom machine vision solutions.
Machine vision strives to automate tasks based on vision. The system consists of a camera and a computer that analyzes images or video. Machine vision is used to both see and understand the view. The system helps identify objects, people or activities. Identification is aided by artificial intelligence and machine learning, which can be used to teach the system to recognize desired things from an image.
Until a few years ago, the benefits achieved with machine vision were marginal. However, technology has advanced tremendously over the last ten years. This is mainly due to rapid development of machine learning methods. The most typical use cases are automatic inspection, object detection, event identification, process control and guidance.
Machine vision is able to detect multiple properties simultaneously and make quick decisions based on the observations.
The machine is more accurate and faster than a human to detect quality deviations. Quality can be monitored around the clock.
Automation brings significant cost savings and streamlines operations. Machine vision assists people and frees up time for more important tasks.
Quality deviations are more reliably detected. More consistent quality improves customer satisfaction and the company's image.
By combining the physical and digital worlds, you create added value for your business. At Empirica, we help you find the best solutions for each case.
We develop custom solutions that utilize machine vision. We also supply all the necessary devices with competitive prices - fast.
We are available to our customers not only during the project but also after the implementation. We take care of the functionality of the services and help users in the case any problems arise.
The machine vision application monitors the outdoor activities of elderly patients and makes alarms when needed. Automated monitoring frees up working time for more important tasks and increases outdoor safety. The application is in test use at the Helsinki City Service Center Kustaankartano.
The purpose of the development work was to test the solution and algorithms developed for Empirica’s pattern recognition and to teach them to identify possible crashes or exits. The system detects crashes outdoors and alerts staff.
NVIDIA is known for its high-performance graphics processors. Thanks to efficient edge calculation devices, lower latency is achieved and the calculation can be performed even with poor network connections.
Intel offers the widest range of devices in the industry for harnessing artificial intelligence, from cameras to edge computing devices. We make extensive use of Intel devices and technologies such as RealSense and Up Squared in our development.
Azure offers a number of services and interfaces that accelerate the development of smart applications. In machine vision applications, we make special use of Azure's Cognitive Services, which have been developed for example for face recognition and image and video analysis.
OpenCV is an open source library originally developed by Intel that provides a unified infrastructure for machine vision applications and accelerates the use of machine vision in commercial products.
Pytorch is another open source library we use to develop machine learning applications. PyTorch enables fast and agile development and efficient production.
The third open source library we use is TensorFlow, developed by Google, which is especially suitable for teaching and utilizing neural network models.
