Obviousness

The following analysis of obviousness for US Patent 11252325 under 35 U.S.C. § 103 considers combinations of prior art references identified in the patent's background section and the motivation a person having ordinary skill in the art (POSITA) would have to combine them.

The key features of US11252325, as described in its abstract and embodiments, include:

1. A computerized method for computing photo quality in a device image acquisition system. [cite: "Abstract", "Embodiment 1"]
2. On-board combining of a plurality of quality indicators (QIs). [cite: "Abstract", "Embodiment 1"]
3. QIs computed from the captured image and its previous image frames. [cite: "Abstract", "Embodiment 1"]
4. A confidence level for at least one of the QIs. [cite: "Abstract", "Embodiment 1"]
5. Using a processor to determine, based on said combining, whether photo quality is acceptable. [cite: "Abstract", "Embodiment 1"]
6. Taking differential action depending on whether quality is or is not acceptable. [cite: "Abstract", "Embodiment 1"]
7. Notably, the patent emphasizes that the weight of one indicator "will take into account data from other quality indicator/s e.g. their quality indicator value, weight, confidence level... and their previous value, weight and confidence level." [cite: "SUMMARY"]

Prior Art References Considered

The patent lists the following as conventional technology constituting background:

• US 20130155474 (referred to as 'a'): Describes providing feedback to a user of a mobile device prior to image capture, based on measured parameters. These parameters can be combined into an "overall quality score" that must exceed a "defined threshold value before the image can be captured." It also suggests feedback like "hold the camera steadier" for motion blur. [cite: "a. US 20130155474"]
• WO 2006040761/US20070195174 (referred to as 'b'): Discloses a system where an interface module allows defining "scene dynamics" (e.g., image motion speed, subject motion speed) and "setting the captured image attributes relative weight for the compution of the total image grade." [cite: "b. WO 2006040761/US20070195174"]
• US20090278958 (referred to as 'c'): Describes that "The scoring of a current base image may be based on scores which have been given to previously captured base images" to avoid redundant calculations. [cite: "c. US20090278958"]

Obviousness Combination: US 20130155474 + WO 2006040761/US20070195174 + US20090278958

A person having ordinary skill in the art (POSITA) would have been motivated to combine these prior art references to arrive at the claimed invention, particularly by enhancing the robustness and accuracy of real-time image quality assessment.

Teaching of References:

• US 20130155474 provides a core system for real-time image quality assessment on a mobile device, combining multiple quality parameters into an overall score, using a threshold to determine acceptability, and taking differential action by controlling image capture or providing user feedback (e.g., for blur due to camera motion). This teaches elements 1, 2, 5, and 6 of the claimed invention. [cite: "a. US 20130155474"]
• US20090278958 teaches computing scores for a current image based on "scores which have been given to previously captured base images." This directly addresses element 3: using QIs computed from "previous image frames." [cite: "c. US20090278958"]
• WO 2006040761/US20070195174 introduces the concept of dynamically adjusting "relative weight" for different image attributes (QIs) based on "scene dynamics" like motion speed when computing a total image grade. This provides a clear teaching of dynamically weighting QIs in a combined score. [cite: "b. WO 2006040761/US20070195174"]

Motivation to Combine:
A POSITA, seeking to improve the photo quality assessment and feedback system of US 20130155474, would find clear motivation to incorporate the teachings of the other references:

1. US 20130155474 + US20090278958: To make the quality assessment more stable and accurate over time, a POSITA would be motivated to use historical data. US20090278958 explicitly teaches basing current image scores on "previously captured base images" to avoid redundant calculations and provide continuity. [cite: "c. US20090278958"] Integrating this into US 20130155474's system would allow for a more consistent and refined "overall quality score" and feedback, leveraging temporal information inherent in a continuous image stream (frames). [cite: "a. US 20130155474"]
2. Adding WO 2006040761/US20070195174: A POSITA would recognize that not all quality issues are equally important in every scene. The dynamic weighting taught by WO 2006040761/US20070195174, which adjusts "relative weight" based on "scene dynamics" (e.g., motion), offers a way to prioritize certain QIs. [cite: "b. WO 2006040761/US20070195174"] For instance, in a scene with significant motion, the blur quality indicator's weight could be increased, aligning the system's assessment more closely with user expectations for that specific scene. This makes the combined system of US 20130155474 and US20090278958 more intelligent and adaptable.

Addressing the "Confidence Level" and Dynamic Weighting:
The primary distinguishing feature highlighted by US11252325 is the inclusion of a "confidence level" for QIs and the dynamic adjustment of weights based on these confidence levels and other QIs' states. [cite: "SUMMARY"]

However, a POSITA would find it obvious to integrate a measure of reliability or "confidence" into the combined system. The description of US11252325 itself states that "all sensors give out errors" and discusses how "recognition or pattern algorithms have assumptions that can be related to 'probability' of the feature been searched." [cite: "Errors and Probability in Quality Indicators"] This acknowledges that assessing the reliability of input data or algorithm outputs is a known problem.

Given that WO 2006040761/US20070195174 already teaches dynamically adjusting weights based on "scene dynamics" (which are effectively other quality indicators like motion speed) [cite: "b. WO 2006040761/US20070195174"], it would be a logical and obvious engineering step for a POSITA to incorporate the reliability or confidence in these "scene dynamics" measurements into the weighting scheme. For example, if the measurement of "image motion speed" (a scene dynamic) is inherently unreliable (e.g., due to low light or complex patterns), a POSITA would naturally assign a lower "confidence" to that measurement and reduce its influence, or the influence of QIs dependent on it, in the overall quality calculation. This directly leads to the claimed feature that "the weight of one indicator will take into account data from other quality indicator/s e.g. their ... confidence level." [cite: "SUMMARY"]

The patent also describes confidence in the context of inter-sensor consistency, such as when "GPS data say location has changed 3 meters in the last 15 milliseconds, but accelerometer data shows no change... then the GPS data may not be trusted, and confidence in the GPS QI for movement, is low." [cite: "Confidence Level of Quality Indicator"] Cross-referencing disparate sensor data to enhance the reliability of individual measurements is a standard technique in sensor fusion and would be obvious to a POSITA seeking to improve the accuracy of any multi-sensor system.

Therefore, the combination of US 20130155474, US20090278958, and WO 2006040761/US20070195174 would lead a POSITA to a system that combines multiple quality indicators, uses previous frames, and dynamically weights these indicators. The further step of introducing a "confidence level" for these indicators and using it to refine the dynamic weighting, particularly in light of inherent sensor errors and algorithmic uncertainties, would be an obvious improvement to enhance the system's accuracy and robustness.