Image sensors for generating floating point numbers

1. An image sensor, comprising:

a plurality of image sensor pixels;

an output line that is coupled to the plurality of image sensor pixels; and

readout circuitry that is coupled to the output line and that converts an analog pixel signal received from a selected image sensor pixel in the plurality of image sensor pixels via the output line into a corresponding floating point number, wherein converting the analog pixel signal into the corresponding floating point number involves performing an exponent conversion and a mantissa conversion, and wherein the readout circuitry uses a first ramp step while performing the exponent conversion and a second ramp step that is different from the first ramp step during the mantissa conversion.

2. The image sensor defined in claim 1 , wherein the readout circuitry comprises:

an analog-to-digital converter (ADC) that generates an exponent value for the floating point number during a first time period and that generates a mantissa value for the floating point number during a second time period after the first time period.

3. The image sensor defined in claim 2 , wherein the readout circuitry further comprises:

an exponent memory circuit that stores the exponent value; and

a mantissa memory circuit that stores the mantissa value.

4. The image sensor defined in claim 2 , wherein the readout circuitry further comprises:

a digital-to-analog converter (DAC) that receives an output signal from the ADC and that supplies an adjustable reference voltage to the ADC.

5. The image sensor defined in claim 4 , wherein DAC supplies a first reference voltage to the ADC when the exponent has a first value and supplies a second reference voltage that is different than the first reference voltage to the ADC when the exponent has a second value that is different than the first value.

6. The image sensor defined in claim 4 , wherein the DAC supplies a nominal reference voltage to the ADC during the first time period, and wherein the DAC adjusts the reference voltage that is supplied to the ADC during the second time period based on the exponent value.

7. The image sensor defined in claim 4 , wherein the ADC comprises a single slope analog-to-digital converter, and wherein the DAC includes binary-weight capacitors.

8. A method of operating an image sensor, comprising:

selecting an image sensor pixel from a plurality of image sensor pixels for readout;

using the selected image sensor pixel to output a pixel signal onto an output line; and

with analog-to-digital converter (ADC) circuitry, converting the pixel signal into a corresponding floating point number by performing an exponent conversion and a mantissa conversion for the floating point number, wherein the ADC circuitry comprises a single slope analog-to-digital converter, and wherein converting the pixel signal comprises using a first ramp signal with a first ramp step during the exponent conversion and using a second ramp signal with a second ramp step that is different than the first ramp step during the mantissa conversion.

9. The method defined in claim 8 , wherein the exponent conversion is performed before the mantissa conversion.

10. The method defined in claim 8 , wherein converting the pixel signal comprises performing data conversion using a fixed nominal reference voltage during the exponent conversion and using an adjustable reference voltage that is no greater than the fixed nominal reference voltage during the mantissa conversion.

11. The method defined in claim 8 , wherein converting the pixel signal comprises performing data conversion using a fixed nominal reference voltage during the exponent conversion and using an adjustable reference voltage that is less than the fixed nominal reference voltage during the mantissa conversion.

12. The method defined in claim 8 , wherein converting the pixel signal comprises performing data conversion using a fixed nominal reference voltage during the exponent conversion and using an adjustable reference voltage that is that is decremented from the fixed nominal reference voltage by some power of two during the mantissa conversion.

13. The method defined in claim 8 , wherein performing the exponent conversion comprises performing the exponent conversion during a first time period, and wherein performing the mantissa conversion comprises performing the mantissa conversion during a second time period that is longer than the first time period.

14. A system, comprising:

a central processing unit;

memory;

input-output circuitry; and

an imaging device, wherein the imaging device comprises:

an array of image sensor pixels;

a column output line that is coupled to a portion of the array of image sensor pixels; and

column readout circuitry coupled to the column output line, wherein the column readout circuitry is configured to perform exponent conversion and mantissa conversion on analog pixel signals received at the column readout circuitry via the column output line, wherein the column readout circuitry comprises:

a digital-to-analog converter (DAC) that generates a fixed reference voltage during the exponent conversion and that generates an adjustable reference voltage that is decremented from the fixed reference voltage by some power of two during the mantissa conversion.

15. The system defined in claim 14 , wherein the column readout circuitry comprises a digital-to-analog converter (DAC) that generates an adjustable reference voltage.

16. The system defined in claim 14 , wherein the column readout circuitry comprises a single slope analog-to-digital converter (ADC) circuit.

17. The system defined in claim 14 , wherein the column readout circuitry has an effective resolution, and wherein ADC circuit has a resolution that is less than the effective resolution.