# US 2026/0095683 A1 · Sparse 4c2+ phase detection auto focus and correlated multiple sampling

> Agents: this data is available as JSON from the patentgrep API and MCP server — see https://patentgrep.com/llms.txt

- Number: US 2026/0095683 A1
- Document ID: US20260095683A1
- Status: Published application
- Kind: A1
- Application number: 19/364,964
- Filed: 2025-10-21
- Published: 2026-04-02
- Earliest claimed: 2023-05-23
- Assignees: OmniVision Technologies, Inc.
- Applicants: Omnivision Technologies, Inc.
- Inventors: Rui Wang
- Web page: https://patentgrep.com/patent/US20260095683A1

## CPC

- H04N 25/78: Readout circuits for addressed sensors, e.g. output amplifiers or A/D converters
- H04N 23/672: {based on the phase difference signals}
- H04N 25/704: Pixels specially adapted for focusing, e.g. phase difference pixel sets
- H04N 25/75: Circuitry for providing, modifying or processing image signals from the pixel array
- H04N 25/76: Addressed sensors, e.g. MOS or CMOS sensors
- H04N: PICTORIAL COMMUNICATION, e.g. TELEVISION

## Abstract

An arithmetic logic unit includes a GC to binary stage, an adder stage, an adder output stage, an adder input latch stage coupled to latch outputs of the GC to binary stage, a feedback multiplexer stage coupled to receive the outputs of the GC to binary stage, a latch output multiplexer coupled to receive outputs of the first adder input latches, where the latch output multiplexer is configured to multiply the outputs of the first adder input latches by either −1 or −2, and an adder input multiplexer stage, where first inputs of the adder input multiplexer stage are coupled to receive outputs of the latch output multiplexer and second inputs of the adder input multiplexer stage are coupled to receive outputs of the second adder input latches. The arithmetic logic performs adaptive correlated multiple sampling for image sensing pixels and phase detection auto focus for other pixels.

## Claims

### Claim 1

A method of operating an arithmetic logic unit (ALU), the method comprising:
configuring a Gray code to binary (G2B) stage to output a reset signal;
configuring first adder input latches of an adder input latch stage to latch the reset signal in response to a first adder input latch enable signal;
configuring the G2B stage to output a first data signal; and
upon determining that the ALU is coupled to an image sensing pixel and that the first data signal is not saturated—
configuring a latch output multiplexer to multiply outputs of the first adder input latches by −2,
configuring an adder stage to sum outputs of the latch output multiplexer and the first data signal from the G2B stage,
configuring second adder input latches of the adder input latch stage to latch outputs of the adder stage,
configuring the G2B stage to output a second data signal,
configuring the adder stage to sum outputs of the second adder input latches and the second data signal from the G2B stage, and
configuring an adder output stage to divide outputs of the adder stage by 2.

### Claim 2 (depends on 1)

The method of claim 1, further comprising, upon determining that the ALU is coupled to an image sensing pixel and that the first data signal is not saturated, configuring second data latches to latch outputs of the adder output stage in response to a second data latch enable signal.

### Claim 3 (depends on 1)

The method of claim 1, further comprising, upon determining that the ALU is coupled to an image sensing pixel and that the first data signal is saturated—
configuring the G2B stage to output a second data signal;
configuring the latch output multiplexer to multiply outputs of the first adder input latches by −1; and
configuring the adder stage to sum outputs of the latch output multiplexer and the second data signal from the G2B stage.

### Claim 4 (depends on 3)

The method of claim 3, further comprising, upon determining that the ALU is coupled to an image sensing pixel and that the first data signal is saturated, configuring second data latches to latch outputs of the adder output stage in response to a second data latch enable signal.

### Claim 5 (depends on 1)

The method of claim 1, further comprising, upon determining that the ALU is coupled to a phase detection auto focus (PDAF) pixel and that the first data signal is not saturated—
configuring the latch output multiplexer to multiply outputs of the first adder input latches by −1;
configuring the adder stage to sum outputs of the latch output multiplexer and the first data signal from the G2B stage;
configuring first data latches to latch outputs of the adder stage in response to a first data latch enable signal;
1 configuring the G2B stage to output a second data signal;
configuring the adder stage to sum outputs of the latch output multiplexer and the second data signal from the G2B stage; and
configuring second data latches to latch outputs of the adder stage in response to a second data latch enable signal.

### Claim 6 (depends on 1)

The method of claim 1, further comprising, upon determining that the ALU is coupled to a phase detection auto focus (PDAF) pixel and that the first data signal is saturated—
configuring the latch output multiplexer to multiply outputs of the first adder input latches by −1;
configuring first data latches to latch a PDAF saturation flag;
configuring the G2B stage to output a second data signal;
configuring the adder stage to sum outputs of the latch output multiplexer and the second data signal from the G2B stage; and
configuring second data latches to latch outputs of the adder stage in response to a second data latch enable signal.

### Claim 7

A method of operating an arithmetic logic unit (ALU), the method comprising:
configuring a Gray code to binary (G2B) stage to output a first reset signal;
configuring second adder input latches of an adder input latch stage to latch the first reset signal in response to a second adder input latch enable signal;
configuring the G2B stage to output a second reset signal;
configuring an adder stage to sum outputs of the second adder input latches and the second reset signal from the G2B stage;
configuring an adder output stage to divide outputs of the adder stage by 2;
configuring first adder input latches of the adder input latch stage to latch the outputs of the adder output stage;
configuring the G2B stage to output a first data signal; and
upon determining that the ALU is coupled to an image sensing pixel and that the first data signal is not saturated—
configuring a latch output multiplexer to multiply outputs of the first adder input latches by −2,
configuring the adder stage to sum the outputs of the latch output multiplexer and the first data signal from the G2B stage,
configuring the second adder input latches to latch the outputs of the adder stage in response to the second adder input latch enable signal,
configuring the G2B stage to output a second data signal,
configuring the adder stage to sum outputs of the second adder input latches and the second data signal from the G2B stage, and
configuring the adder output stage to divide outputs of the adder stage by 2.

### Claim 8 (depends on 7)

The method of claim 7, further comprising, upon determining that the ALU is coupled to an image sensing pixel and that the first data signal is not saturated, configuring second data latches to latch outputs of the adder output stage in response to a second data latch enable signal.

### Claim 9 (depends on 7)

The method of claim 7, further comprising, upon determining that the ALU is coupled to an image sensing pixel and that the first data signal is saturated—
configuring the latch output multiplexer to multiply outputs of the first adder input latches by −1;
configuring the G2B stage to output a second data signal; and
configuring the adder stage to sum the outputs of the first adder input latches and the first data signal from the G2B stage.

### Claim 10 (depends on 9)

The method of claim 9, further comprising, upon determining that the ALU is coupled to an image sensing pixel and that the first data signal is saturated, configuring second data latches to latch outputs of the adder stage in response to a second data latch enable signal.

### Claim 11 (depends on 7)

The method of claim 7, further comprising, upon determining that the ALU is coupled to a phase detection auto focus (PDAF) pixel and that the first data signal is not saturated—
configuring the latch output multiplexer to multiply outputs of the first adder input latches by −1;
configuring the adder stage to sum outputs of the latch output multiplexer and the first data signal from the G2B stage;
configuring first data latches to latch outputs of the adder stage in response to a first data latch enable signal;
configuring the G2B stage to output a second data signal;
configuring the adder stage to sum the outputs of the latch output multiplexer and the second data signal from the G2B stage; and
configuring second data latches to latch outputs of the adder stage in response to a second data latch enable signal.

### Claim 12 (depends on 7)

The method of claim 7, further comprising, upon determining that the ALU is coupled to a phase detection auto focus (PDAF) pixel and that the first data signal is saturated—
configuring the latch output multiplexer to multiply outputs of the first adder input latches by −1;
configuring first data latches to latch a PDAF saturation flag;
configuring the G2B stage to output a second data signal;
configuring the adder stage to sum outputs of the latch output multiplexer and the second data signal from the G2B stage; and
configuring second data latches to latch outputs of the adder stage in response to a second data latch enable signal.

### Claim 13 (depends on 7)

The method of claim 7, further comprising configuring the first adder input latches to latch an average of the first reset signal and the second reset signal.

### Claim 14 (depends on 7)

The method of claim 7, further comprising configuring a PDAF circuit of the ALU to determine whether the ALU is coupled to the image sensing pixel or a phase detection auto focus (PDAF) pixel.

### Claim 15

A method of operating an arithmetic logic unit (ALU), the method comprising:
receiving a comparator output generated in response to a comparison of an analog signal from a column bitline to a ramp signal;
generating a front end latch enable signal in response to a falling edge in the comparator output;
latching Gray code outputs from a Gray code generator in a front end latch stage in response to the front end latch enable signal;
latching outputs of the front end latch stage in a signal latch stage in response to a signal latch enable signal;
converting the latched Gray code outputs to a binary representation using a Gray code to binary (G2B) stage;
determining whether the ALU is coupled to an image sensing pixel or a phase detection auto focus (PDAF) pixel; and
selectively performing one of a correlated multiple sampling operation and a phase detection auto focus operation based at least in part on the determination.

### Claim 16 (depends on 15)

The method of claim 15, further comprising determining, using a saturation circuit of the ALU, whether a data signal is saturated.

### Claim 17 (depends on 16)

The method of claim 16, wherein selectively performing the correlated multiple sampling operation comprises, upon determining that the ALU is coupled to the image sensing pixel and that the data signal is not saturated:
configuring a latch output multiplexer to multiply outputs of first adder input latches by −2;
configuring an adder stage to sum outputs of the latch output multiplexer and a first data signal from the G2B stage;
configuring second adder input latches to latch outputs of the adder stage;
configuring the adder stage to sum outputs of the second adder input latches and a second data signal from the G2B stage; and
configuring an adder output stage to divide outputs of the adder stage by 2.

### Claim 18 (depends on 16)

The method of claim 16, wherein selectively performing the phase detection auto focus operation comprises, upon determining that the ALU is coupled to the PDAF pixel and that the data signal is not saturated:
configuring a latch output multiplexer to multiply outputs of first adder input latches by −1;
configuring an adder stage to sum outputs of the latch output multiplexer and a first data signal from the G2B stage;
configuring first data latches to latch outputs of the adder stage;
configuring the adder stage to sum outputs of the latch output multiplexer and a second data signal from the G2B stage; and
configuring second data latches to latch outputs of the adder stage.

### Claim 19 (depends on 15)

The method of claim 15, wherein latching the Gray code outputs from the Gray code generator includes latching the Gray code outputs from a Gray code generator that is shared among a plurality of column ALUs of an image sensor readout circuit.

### Claim 20 (depends on 15)

The method of claim 15, further comprising outputting digital image data signals or phase detection signals from the ALU through output switches to a shift register readout.

### Claim 21 (depends on 15)

The method of claim 15, wherein determining whether the ALU is coupled to the image sensing pixel or the PDAF pixel includes determining whether the ALU is coupled to the imaging sensing pixel or the PDAF pixel using a PDAF circuit of the ALU.

### Claim 22 (depends on 15)

The method of claim 15, further comprising:
providing outputs of an adder output stage of the ALU to first inputs of a feedback multiplexer stage of the ALU; and
providing outputs of the G2B stage to second inputs of the feedback multiplexer stage.

### Claim 23 (depends on 1)

The method of claim 1, further comprising configuring a front end latch stage of the ALU to latch Gray code outputs from a Gray code generator in response to a comparator output.

### Claim 24 (depends on 1)

The method of claim 1, further comprising configuring a phase detection auto focus (PDAF) circuit of the ALU to determine whether the ALU is coupled to the image sensing pixel or a PDAF pixel.

### Claim 25 (depends on 1)

The method of claim 1, further comprising configuring a saturation circuit of the ALU to determine whether the first data signal is saturated.
