EYEphone
Alexandra Perron, Loïc Hermant, Régine Roncucci et Léon Grillet
A sensors project
The EYEphone project is a one week teamwork proposal based on interdisciplinary knowledge, involving biology, physics and informatics.
“Does the eye adapt faster to a change in light exposition than a phone camera? “
For this project, we dedicated our work to the study of two differents sensors, a biological and an electrical one. We chose the human eyes as our biological sensor. The Human eye is constantly adapting to the changes of light, even if we can’t always realize it. Our brain allows us to interpret changes of light in less than a fraction of second and adapt the size of the pupil by contracting or dillatating the iris to obtain a good image resolutions.
The second sensor that we study is the camera of a phone (Sony Xperia 5). Camera can adapt autonomously to light using algorythm to equilibrate the amount of black and white pixels that are in the same frame.
Our experiment is divided in 3 main parts : Collecting the data, analysing the data and finally interpreting the results.
Collecting the data
We created a set up to take video of the eye of our subjects while exposed to light.
The experiment consists in changing the light received by an eye, projecting a flash on it and at the same time filming the change of pupil size. We ask to 10 persons to test the experiment. People during the experiment weren’t allowed to move and the installation was made for being sure that people put their heads in the same place.
The flash was activated at a distance of 23 cm to the eye of the sample eye and during 3 seconds and then we let the eye reposed for more or less one minute.
We made 3 replicates for each person to obtained quantitative data.
We followed exactly the same protocol for the phone camera. The camera that was filming has been exposed to the flash and we recorded its focus. We repeated the experiment 15 times.
The analyse of the data
We extracted 5 pictures of each video that we take. Those pictures were always taked based on the time of adaptation. The time between two pictures is equal to the time of adaption divided by 5.
To analysing our experimental data, we used the imageJ software that allows us to analyse the color density of our pictures.
We use this software to extract from the images the length of the pupil in order to see the difference from the size of the pupil before, during and after the exposition to flash. We study also the adaptation time of the pupil to move from the dilated phase to the contrationnary phase of the pupil. Thanks to imageJ, we manage to analyse our picture putting them in black and white and quantifying were is the highest concentration of black pixels that corrspond to the pupil.
As it regards the phone camera, we made almost the same thing, but this time, we measure the lenght of the white spot from the flash.
Interpretation of results
To be able to interpret our results correctly, we normalized our experimental data. Thanks to this method, we have values which have for maximum 1 and thus we can try to interpret the evolution of curves.
Our results were represented in this graph. We observed that the curve of the phone camera decreases very quickly from 1 to 0.4 in 0.6 seconds then slowly in 0.4 in 0.3 until 1 second whereas for the curve of eyes, we observe that the curve decreases slowly and almost linearly in 1 in 0.7 in 1 second.
From these results, we can say that the phone camera adapts itself more quickly than eyes to the light. We also understand that the evolution of adaptation of the phone camera is in two steps which are to get the image and after use algorithm to focus on the details of the image.
Critical analysis of the project
To conclude, we have an idea of evolution of eyes and phone camera adaptation. However, we have not taken enough paramaters to compare exactly eyes and phone camera. We need to take into account others paramaters such as the retinal adaptation. To improve this experiment, we need to reduce experimental and analysis uncertainties of our experiments. To obtained more relevant data, we should also study a bigger sample. We need also to automatize data analysis because it take too much time and there is more chance to realize mistakes.
If you want to know more about eyes and phone camera adaptation, here some links:
Video:
Article:
Check us on Twitter: https://twitter.com/1995Regine?lang=fr
No comments:
Post a Comment