In Part 2, we will explore the new Surface OCR tool.
Optical Character Recognition (OCR)
The Surface OCR tool allows you to recognize and extract a string of text from surfaces, using either 3D height map scan data or 2D intensity scan data.
Surface OCR adds a valuable capability for inspection applications that decode flat or embossed characters in markets such as automotive (stamped parts), battery and packaging (track and trace), and rubber & tire sidewall (DOT codes).
How It Works
Gocator® scans the target and generates a 3D surface. Next, simply drag and drop the OCR tool, set the region of interest, and decode characters in the scan data. The OCR tool is based on deep learning and comes with a pre-trained dataset, eliminating the need to perform training on your own.
Surface OCR works on both 2D intensity data and 3D height map data. While 2D recognition is an important capability, the ability to read characters in 3D is unique to Gocator. 3D OCR is essential in many applications, such as tire sidewall DOT-code reading, where the characters are embossed.
Toggling Search Parameters, Blacklist/Whitelist Mode, and Inversion
With the Surface OCR tool, you also have the ability to toggle between Alphanumeric or Numeric search parameters, depending on your scan target.
Blacklist and Whitelist modes allow you to exclude and include specific characters, respectively. This gives you additional control, keeping the OCR process simple and robust, and resulting in more repeatable results. You can even use the Invert function to swap intensity values in the data the tool uses to perform OCR, which is useful when you need to read and recognize light text on a dark background.