Color for the Colorless

By: Ibrahim H. & Caleb Sh.
Year: 2020
School: Talbert Middle School, 8th grade
Division: Junior

For this project, we designed a color sensor that is capable of distinguishing different colors in its environment to help people who are visually impaired. To do this we used a color sensor which works by measuring the amount of reflected light and filtering each color in the light. It consists of photodiodes that do the job of measuring 3 colors, red, green, and blue(RGB). Each of these colors can be mixed together to obtain every color in the color spectrum.To use the color sensor, it must be properly calibrated and code must be accurate to be able to carry out its purpose. The color sensor uses a portable display to communicate with and display the color it is seeing. For this experiment, we measured the maximum distance for which the color sensor could be used and the accuracy for detecting each color.

Using a combined, but limited budget, we were able to create a color sensor to accurately depict colors. We each researched different parts of the project to help us better understand the way it should be carried out. Some of the research conducted included the programming language being utilized, how each port between board and the color sensor should be connected, and the overall cost of the project. Continuous research gained us knowledge on the different devices that were applicable to our needs. One of these was the Arduino Uno board which was both inexpensive and functional. Along the way, this device needed a way to be interacted with in order to become usable without a large device like a computer. Using a touch sensor, the color sensor was able to run when the user pressed a button, rather than running whenever the device is powered on.needed a way to be interacted with in order to become usable without a large device like acomputer. Using a touch sensor, the color sensor was able to run when the user pressed a button,rather than running whenever the device is powered on.The materials we used include the Ardiuno board, a color sensor TC230 TC3200,capacitive touch sensors, a portable battery or laptop, colored objects, and a measuring tool. Toperform this experiment, we used the code library I created on the Arduino website and uploadedit to the device. Afterwards we made sure the ports between all the devices were properlysituated, and then powered on the device and allowed it to read different colors. Some of theinformation that was recorded include the distance, the color of the object, the color returned bythe sensor, and the optimal height for each distance. The colors that we tested were red, green,and blue because the computer functions by sensing each of these colors and then combiningthem to see what it produces.The graphs and data show data that shows the accuracy of the color sensor detecting red,green, and blue. By detecting these three colors, it can be used to determine every other color inthe rgb color table. These colors are digitally combined to return the color it is really seeing. Onaverage, the graphs show that the optimal distance that the color sensor shows full accuracy is 23cm. The optimal height is eyeballed and can be easily applied by using the sensor when the lightis centered on the object. Some of the disadvantages that were detected during the experimentinclude the large area that the sensor uses to measure reflected light. This makes it difficult todetect objects with a mix of colors, but the area decreases as the color sensor gets closer to thetarget. The advantage of this is an increase of accuracy for detected colors on largest objects. Thedata is from a mathematical formula that gives the square of the light frequency. This gives theneeded wavelengths for the arduino to detect colors in the visible color spectrum. Moremathematics was used to get the accuracy. The distance from the rgb of the color of the paperand the amount of times the correct color was given. Towards the end of the results the accuracyof the color sensor started to decrease at a rate of 10% that showed it was coming close to its