Colors Detector with Real Human Voice Pronouncing
The aim of creation of this color detector device is to help blind people, or people with a problem of daltonism, who differ colors with difficulties or wrongly. This device analyzes the color of a surface. When it determines the color, it pronounces from the speaker.
The device is portable and consists of main parts : Three color sensors (TRS17xx), using primary colors (red, green, blue) and three LEDs (for lightening of a surface) placing inside the same IC; Text-to-Speech IC WTS701E from Winbond with SPI protocol operated in 3,3V; audio filters, one audio amplifier LM386 and Renesas MCU R8C/13 as main processor.
There are 40 texts of the names of colors in this device. These texts are in English language. To detect the correct color it is utilizing above-stated algorithm using original database of 40 colors (storied in internal Data FLASH). That vector with most small value in the indicated above figure is between points of green and red dots. The color that we need to recognize from surface is the most eligible to the color applicable to the green dot.
Colors Detector
Project Firmware, Schematic and Documentation
circuitcellar.com
