The Scorpion Vision Bacon Inspector

We eat a massive amount of bacon in this country. In fact, an online resource stated that Bacon is an industry worth over ￡900 million with consumers currently purchasing over 150,000 tonnes per year. Rashers remain as the most popular form of bacon (SOURCE).

In addition to the regular fry-ups we have at home, we also order our fast food with bacon. A famous fast food brand sells over 50% of its products with bacon. Hence, this must translate to millions and millions of bacon rashers being consumed each week. 

Do you ever wonder how supply keeps up with demand? Or perhaps even more importantly, have you ever noticed how the bacon on your bacon burger is consistently of the same colour, shape, and size? How come we never see burnt bacon? I burn bacon at home all the time! 

Well, we can answer some of those questions as we have provided a Scorpion Vision high-speed bacon rasher inspection and counting system for the industry.

16 Cameras Simultaneously Looking at 1300 Rashers of Bacon Per Minute

The inspection system uses 16 individual Scorpion Stinger 2D cameras to inspect 16 lines of bacon rashers moving on a conveyor that is travelling at 14 metres per minute. That literally translates to 1300 rashers of bacon per minute or nearly 500,000 bacon rashers in 24 hours from one factory. 

Our camera-based system actually looks at every individual bacon rasher and measures it to ensure that it meets the criteria set by the manufacturer. If it does meet the criteria, the rasher is then allowed to drop into a container along with a pre-set number of other rashers that have also passed the inspection. The vision system knows when the container is full as it has counted everything that goes into it. At this point, the container is released to be sealed and sent off to be dispatched. 

If, on the other hand, the rasher does not pass the inspection, the vision system identifies it for rejection. Its position on the conveyor belt is recorded and the item is tracked to the end of the conveyor where it is pushed off the conveyor by an air jet.

It's More Complex Than It Sounds

The clever bit here is the Scorpion Vision bacon counting system as it has to manage a large and complex number of scenarios: 

All potential rashers or objects that are assessed as potential rashers by Scorpion Vision are assessed. The count analysis includes a top, left, bottom, and right extent, which is generally different from that defined for measurement analyses. If any rasher touches these boundaries then it is assessed as touching an edge. 

Potential rashers that touch the edge of the image are assessed. If the region of interest being processed includes the edge of the belt, then reflections of lights on the edge of the belt may result in false potential rashers, which are also assessed. Therefore, it is critical that the region of interest defined for analyses excludes any section of the belt which may give rise to false potential rashers. 

For the customer, it is critical that the number of rashers in each box is correct, with no under-counts and a maximum of 1 over-count. Rashers may be seen in more than one image, thus it is essential to have the option to assess whether a rasher being analysed in the current image is the same rasher as one that was analysed in the previous image or a different one.

Massive Production Gains by Minimising 'Giveaway'

The gain for this bacon producing factory is significant because in the past, bacon was placed into the containers by hand and its weight was roughly measured. This resulted in what the industry calls ‘giveaway’. So, the move to a system that counted individual rashers, whilst also measuring them to improve quality, makes complete sense. Above all, it saves money as well!

And as ever, here's a brief video. Watch out for the small bits of bacon being pushed off the end of the conveyor by the air jet.