Sophisticated technologies define modern food production. Simultaneously, improving quality, safety, and security remain critical issues in food industry. Process control and automation regularly appear among the top priorities identified in food engineering. Advanced monitoring and control systems are developed to facilitate automation and flexible food manufacturing.
OptoMotive´s cameras not only improve the quality, but also reduce product waste and increase throughput.
• defect inspection,
• contamination inspection,
• automated filling lines control,
• sorting of bottles/crates in correct sizes and variants,
• hygiene inspection,
• fill level inspection,
• label inspection,
• control of bottle cap placement,
• foreign objects inspection,
• sorting of fruit and vegetables,
• sorting of nuts, coffee beans and rice,
• cookie shape inspection,
• bread dough shape inspection,
• making of cuts on bread,
• measuring the height of chocolates,
• verifying the volume of a block of cheese.
In every automated process, computer vision is needed to automate productions and to improve quality and throughput. Bread making, too, has become a fully automated industrial process in modern factories. One might wonder where to find applications of computer vision in automated bread baking. But there are actually numerous examples.
One of the examples really stands out for its complexity and actuality. Bread loafs have different cuts on top. Bakeries use their own distinctive pattern to mark their products. This way they know what kind of recipe was used to bake it, and which store it has to go to. Making cuts on the top side of the dough before baking is a simple task for a human, but not so easy for a robot. If the robot wants to cut the dough precisely, it needs the precise trajectory of its knife. But dough position and shape varies.
How can this be dealt with? The best way is to perform a 3D measurement of the dough. For this purpose, a laser line projector, an FPGA-based camera with laser triangulation core, which does image processing and outputs only profiles instead of images, are needed. These profiles are used to generate a 3D model of dough in its real coordinates. Such a 3D model is basis for calculation of robotic knife trajectory. It's as simple as that.
In order remain competitive the rest of the world, crop harvesting in Europe is becoming a fully automated process. How can we make sure that the individual vegetables or pieces of fruit are of good quality and ready to be sold? A single rotten item in a crate or basket can spoil all the others before they reach customers. Computer vision can be employed to solve this problem.
People usually rely on the colour of vegetables to determine their quality. For this application, the colour reproduction of a vision system’s camera is essential. In such a camera, sensors with high resolution and brilliant colours have to be used.
When this camera is FPGA-based, the on-board FPGA contains an image processing engine to distinguish between good and bad vegetables. The camera passes information on the quality of the product to a controller, which throws away rotten ones, or sorts products in different baskets.