Acquire images with a sensor and a microcontroller
Ioan Ciascai, Technical University of Cluj-Napoca, Cluj-Napoca, Romania, and Liliana Ciascai, Babes-Boylai University, Cluj-Napoca, Romania; Edited by Martin Rowe and Fran Granville - September 23, 2010
| View as PDF |
The TAOS (Texas Advanced Optoelectronic Solutions) TSL1412S image sensor, IC2, can acquire a linear image of 1536×1 pixels, or 400 dpi (Figure 1). It uses a single voltage supply, and you can control it with just a few digital signals. Thus, you can design an image-acquisition system that uses the sensor and an AVR ATmega328 microcontroller, IC1.
| Read More |
The conversion time for a 16-MHz-frequency clock is approximately 16 μsec, which corresponds to an integration time of approximately 25 msec. Because the conversion frequency is 1 MHz—higher than that of IC1’s recommended frequency of 200 kHz—you reduce the ADC’s precision from 10 bits to 8 bits. The microcontroller saves a byte for each pixel, which lets you save the data to the microcontroller’s internal memory for one frame. The rest of the microcontroller’s 2-kbyte memory performs stack and data-acquisition tasks.
The system quickly processes the sensor’s analog output signal through the analog comparator of the microcontroller’s internal schematic. You can make a comparison with a fixed voltage using an internal voltage reference of 1.25V and a resistive divider or a variable voltage you can obtain from a DAC or a PWM (pulse-width-modulated) signal the microcontroller’s timer generates.