Utilizing Deep Learning Algorithm and Image Processing Techniques for the Localization and Quantitative Analysis of Infographics in Resumes

Infographics have proven to be more effective than traditional prose in conveying complex subjects. In recent years deep learning approaches have demonstrated their capability in an array of applications, including image identification, which encompasses pattern recognition and the application of artificial intelligence. A variety of infographics are employed in order to gauge skills in resumes. The aim of this study is to identify the existing infographics present in resume, locate the associated text pertaining to each infographic, and subsequently quantify each infographic with numerical values. For the purpose of detecting infographics, the You Only Look Once (YOLO) algorithm was utilized, while the Optical Character Recognition (OCR) was applied for detecting the associated text. The differentiation between the filled and empty sections of the infographic was accomplished through the implementation of image intensity histogram analysis, thresholding, and contouring.