Digitalização da coluna por meio da visão monocular com projeção de luz estruturada

Abstract

According to data from the US National Institute of Occupational Safety and Health (NIOSH), the main health problems found in workers today are pain in the scapular region. This is because many people have a job peculiar to a certain activity, for example, because they work all day in front of the computer, in agricultural machinery operations or even in jobs that require great physical effort. These workers end up spending the entire day with a completely inadequate posture, thus developing diseases and back pain. Another reason that contributes to this condition is the fact that most people do not focus on preventive work, such as postural analysis to assess whether there is any irregularity in their spine. Many choose to practice a work activity that, many times, incorrectly, harm the health of the spine. Lifting excessive loads without proper physical preparation, for example, can trigger various diseases in the region. On the other hand, many health professionals do not use image analysis, as they depend on high-cost software, thus performing only a visual analysis. This leads to inferences in the results according to the degree of knowledge of each professional. With the advancement of technology in favor of academic studies, it was possible to develop a system capable of identifying the 3D profile of the spine in the individual, very similar to the large paid software on the market; however, that of free access, using accessible equipment. The development of this project was based on the principle of structured light which uses the light source through monocular vision, which uses only a webcam, both arranged at angles in relation to each other and to the individual. Several algorithms based on mathematical calculations have been developed to correct the perspective deformations of the observer's relative view in relation to the real shape of the object. We used objects with known heights to carry out the calibration of the experimental arrangement and, afterwards, pictures of the horizontal lines projected on the individuals' backs were taken in such a way that it covered the entire human back. These images were submitted to algorithms, in order to result in a three-dimensional image of the individual's spine. Thus, this project contributes through images in three dimensions, as a complementary tool that health professionals can use to offer a report to their patients in a more reliable way and at an affordable cost.