Bespoke AI technique delivers ‘utopian’ vision of product presentation

Scorpion Vision’s advanced stereo vision and neural network solution, the Scorpion 3D Neural Camera, can elevate the presentation of retail packaged fresh produce to unprecedented levels.

The Scorpion 3D Neural Camera can elevate the presentation of retail packaged fresh produce and reduce the waste associated with inconsistent trimming.

Scorpion Vision is pushing the envelope in terms of what can be achieved by applying AI to machine vision in an automation context. By guaranteeing that processes such as sprout, carrot and leek trimming are consistently completed within the tightest visual tolerances, Scorpion’s bespoke AI approach enables food processors to achieve utopian presentation that will meet the approval of even the most discerning shoppers. The Scorpion 3D Neural Camera has also been heralded as a food waste saviour, as food factories can put a stop to perfectly good food being discarded due to inconsistent and sub-standard inline preparation techniques.

The advent of AI presents an exciting opportunity to enhance the capabilities and performance of vision systems in food processing applications. On the back of this, many off-the-shelf AI-camera solutions have become available.

Paul Wilson, MD of Scorpion Vision Ltd, explains why these ‘plug and play’ cameras can’t match the consistent performance and premium presentation that can be achieved by Scorpion Vision’s 3D technology with bespoke AI: “Lots of companies have jumped onto the AI bandwagon, thinking that all you have to do is programme a camera with deep learning software. It is not that simple, which is why, although off-the-shelf camera solutions appear cheaper at a glance, they rarely provide more than 80% reliability. Integrating AI into an automation system that works 100% of the time requires a deep understanding of the build challenges.”

Scorpion builds AI vision systems from scratch to deliver robust and reliable application-specific solutions to food processing problems.

Paul continues: “We already have our proven automation platform so have been able to build AI into that with great success. Our AI cameras are designed from the ground up to suit the application. That means we can optimise the design and the parameters to enhance performance on that product. If we want to use more powerful lighting or move the lighting outside the camera, for example, we can do, because we have complete control over the build.”

AI-based vegetable processing

Scorpion Vision’s bespoke AI approach can be applied to automated vegetable trimming processes to achieve repeatable precision.

With an organic object such as a vegetable, there is no fixed size, shape or colour, and this presents a problem for classic machine vision camera systems that rely only on known shapes or patterns.

Accurate trimming is key to the presentation of vegetables destined for retail sale. If the leek is cut too short it ‘telescopes’ and dries out. If it is left too long, unsightly roots are left. However, a standard 3D camera will struggle to determine the stem plate when it is obscured by roots or soil. In this scenario, applying AI improves machine vision performance dramatically.

Access to leading technology

The system is based on the 3D Stinger – designed by Tordivel AS to be used in cutting-edge 3D stereo vision applications. Scorpion Vision Ltd can also design automation systems incorporating machine vision technology from other manufacturers. Since signing a distribution agreement for the UK and Ireland with fast-growing global camera and vision systems manufacturer HIKROBOT in 2017, Scorpion Vision Ltd has installed a number of systems based on products from HIKROBOT’s innovative portfolio.

Check Also

Dave Oakes, Powelectrics MD, reviews the first half Of 2022

We caught up with Dave Oakes, Powelectrics MD to find out how the company is …

Fluke helps customers put environmental and financial value on air-compressor maintenance with new tool

Fluke has launched a new CO2 emission indicator to the online LeakQ reporting tool. In addition …