Chocolate Assembly Verification
Identify, count, and verify chocolates based on their appearance for packaging and assembly verification using deep learning software
For certain food packaging applications, manufacturers must rely upon visual inspection to ensure the quality of the final assembly. Visual variations of food products can cause complications for an automated food and beverage inspection system. This is the case for chocolates, which arrive for packaging in differently configured boxes. The inspection system must successfully verify that each spot contains a piece of chocolate and that it is of the correct type.
Holiday-themed (e.g. Christmas or Valentine’s Day) chocolate boxes present a particular challenge. The same chocolates may be present in the boxes, but their position may change based on the box’s theme. The manufacturer must locate each chocolate and verify the correct type is in the correct position.
Similarly, two chocolate boxes might use the same packaging (e.g. box of 6 chocolates) but be filled with different flavors (e.g. milk vs dark chocolate selections). In this case, the manufacturer needs to count the chocolates and verify the correct chocolate selection is present.
Cognex ViDi uses deep learning image analysis software to automate the task of locating and identifying multiple features within a single image. Cognex ViDi can generalize the distinguishing features of various types of chocolate based on their size, shape, and surface features.
With the ViDi Blue-Locate tool, a user can train the tool to locate each type of chocolate that will need to be found. Users can build a database of various types of chocolates for the Blue tool to find, which can then be used for packaging verification.
Once trained, the image can be split into different regions where the Blue-Locate Tool will check for the presence of a chocolate as well as verify that it is of the correct type. Multiple configurations can also be created for situations where a single line may have variations in packaging. In this way, a user can automate the verification of chocolate packaging using just one tool.