Desenvolvimento de algoritmos de analise optica do olho humano em uma plataforma livre
Resumo
Resumo: A visão permite a obtenção de informações a respeito do universo, trazidas aos olhos pela luz. Há muitos séculos, estudiosos elaboram hipóteses sobre o funcionamento desse sentido e os fenômenos luminosos. Desde então, o entendimento da luz evoluiu substancialmente graças a teorias físicas e de engenhosos experimentos com espelhos, lentes, interferômetros e a interação com a matéria. Apesar de ser percebida como um fenômeno emergente de uma partícula quântica, a sua interpretação como uma onda guiada por raios ainda é extremamente útil. Por outro lado, muitos detalhes do funcionamento do olho humano permaneceram ocultos até recentemente, com a sofisticação de equipamentos de medição e o processamento de dados em computadores. Diversos trabalhos buscaram modelar matematicamente o olho utilizando medições reais, mas ainda não há um consenso sobre a melhor forma de representar os elementos ópticos do olho, em particular o cristalino. Esses modelos são de vital importância para os avanços no entendimento da visão e de grande interesse da medicina, como no ramo de cirurgias refrativas. Neste trabalho foi desenvolvida uma simulação do olho humano no GeoGebra, uma plataforma aberta que possibilita a edição e disponibilização das construções feitas nela para o público. Para isso, foram desenvolvidos algoritmos de análise óptica restrita ao um plano que contém o eixo óptico, em que as superfícies são definidas por funções ou conjuntos de pontos, para meios homogêneos e com gradiente de índice de refração (GRIN), de forma que possam ser adaptados para qualquer plataforma com as ferramentas matemáticas necessárias para processá-los. Os algoritmos foram implementados em simulações isoladas, entre as quais está um modelo da estrutura GRIN do cristalino, que reproduz os valores de poder refrativo encontrados na literatura, e uma proposta de tratamento de dados obtidos em medições de superfícies sem interpolação através de equações lineares, cujos resultados convergem para o caso suave quando o número de pontos é da ordem de centenas. As escolhas feitas durante a integração dos algoritmos priorizaram o desempenho computacional do software, de forma que é possível utilizá-lo mesmo em navegadores de internet. Os usuários podem comparar a qualidade óptica do olho humano para diversos parâmetros e atestar o aumento do poder refrativo graças à estrutura GRIN do cristalino Abstract: Vision makes it possible to obtain information about the universe, brought to the eyes by light. For several centuries, scholars have developed hypotheses about the workings of this sense and the optical phenomena. The understanding of light has evolved substantially since then, thanks to physical theories and ingenious experiments with mirrors, emergent phenomenon of a quantum particle, its interpretation as a wave guided by rays is still extremely useful. On the other hand, many details of the functioning of the human eye remained hidden until recently, with the sophistication of measuring equipment and data processing in computers. Several works have sought to mathematically model the eye using real measurements, but there is still no consensus on the best way to represent the optical elements of the eye, particularly the crystalline lens. These models are of vital importance for advances in the understanding of vision and of great interest in medicine, such as in the field of refractive surgery. In this work, a simulation of the human eye was developed in GeoGebra, an open platform that allows editing and publishing of constructions made in it. For this, optical analysis algorithms restricted to a plane that contains the optical axis were developed, in which the surfaces are defined by functions or sets of points, for homogeneous and gradient refractive index (GRIN) media, so that they can be adapted to any platform with the mathematical tools needed to compute them. The algorithms were implemented in isolated simulations, among which is a model of the GRIN structure of the lens, which reproduces the values of refractive power found in the literature, and a proposal for the treatment of real surface data without interpolation through linear equations, whose results converge to the smooth case when the number of points is on the order of hundreds. The choices made during the integration of the algorithms prioritized the computational performance of the software, so that it is possible to use it even in internet browsers. Users can compare the optical quality of the human eye for various parameters and attest to the increase in
refractive power thanks to the GRIN structure of the lens
Collections
- Física [46]