Reading Embossed Fonts on Smartphone Cases Using Deep Learning–Based OCR

The Challenge

Embossed markings on smartphone cases present a significant challenge for conventional Optical Character Recognition (OCR) systems. Variations in font type, character size, material texture, and inhomogeneous backgrounds often result in unreliable reading performance. Traditional OCR methods depend heavily on consistent contrast and uniform surfaces, conditions that are rarely present when dealing with embossed characters on textured materials.

Low contrast between characters and background, combined with textured surfaces, leads to inconsistent results with conventional OCR. In addition, multiple font variants and different character dimensions must be verified to ensure product compliance. The system therefore needed not only to read the characters reliably, but also to check for legibility, completeness, and correct font variant under challenging environmental conditions.

The Solution

To overcome these limitations, a Deep Learning–based OCR solution using senswork’s ViDi Read was implemented. ViDi Read was trained using a comprehensive and diverse dataset that reflected real production variability. The training images included a wide range of lighting situations, background structures, embossing depths, font styles, and character sizes.

Certified System Integrator Program

Set Yourself at the Forefront of the Global Vision Market

Vision system integrators certified by A3 are acknowledged globally throughout the industry as an elite group of accomplished, highly skilled and trusted professionals. You’ll be able to leverage your certification to enhance your competitiveness and expand your opportunities.

GET CERTIFIED

Instead of relying on fixed contrast thresholds or predefined character templates, the system learned to identify relevant visual features directly from the data. This enabled the solution to generalize effectively across variations in topology and illumination. Even when characters appeared with minimal contrast or uneven shading, the trained model was able to accurately detect and interpret them.

Results

This Deep Learning–based OCR solution demonstrated reliable and sturdy inspection performance. It successfully read embossed fonts on inhomogeneous and textured surfaces, maintaining high accuracy even under varying lighting conditions. The system accurately recognized different font styles and character sizes while verifying legibility and completeness of the markings.

Learn more about how Vision AI is used for quality control here.