Higher-resolution cameras usually need more pixels, smaller sensors or more sophisticated lenses. But researchers have demonstrated another approach: extracting information from within the pixels that cameras already have.
The technique could eventually improve imaging in satellites, industrial inspection systems and scientific instruments without requiring a new generation of image sensor chips.
The Hidden Information Inside Every Pixel
A digital camera captures images using millions of tiny light-sensitive elements called pixels. Each pixel measures the light falling on it and produces an electrical signal.
Normally, the camera treats each pixel as a single measurement.
Researchers at the Chinese Academy of Sciences have demonstrated that a pixel may contain more useful spatial information than conventional camera systems extract.
Their technique, called pixel-segmenting imaging, measures how different parts of the same pixel respond to incoming light. It uses specialized optics and an optimization algorithm to map variations in the pixel’s sensitivity.
The researchers then use that information to reconstruct finer image details.
In experiments, they demonstrated a method corresponding to a 4 × 4 grid of sub-pixel regions, effectively reducing the sampling unit to one-sixteenth of a conventional pixel.
Why This Matters for Camera Technology
The conventional way to improve resolution is to pack more pixels onto a sensor.
But smaller pixels collect less light individually, which can create noise and sensitivity problems. Increasing pixel counts also increases the amount of image data that must be processed.
The new approach tackles the problem differently.
Instead of manufacturing more physical pixels, it extracts additional spatial information from the response of existing ones.
In one laboratory comparison, the researchers reported improvements of 2.3 times in a measure of image sharpness and 3.1 times in a structural similarity measure.
These are specific experimental results, not evidence that every camera can suddenly produce images with sixteen times more detail.
What Could Change?
The technology could be particularly useful where replacing an imaging sensor is expensive or difficult.
Satellite imaging systems could potentially benefit from finer image reconstruction. Industrial inspection cameras might identify smaller manufacturing defects, while scientific instruments could extract more information from existing detectors.
There is also a broader implication.
Modern cameras increasingly depend on software to improve image quality. Many AI-based techniques estimate what missing details should look like.
This approach instead uses measured physical properties of the sensor to recover information. That distinction could be important in scientific applications where an image needs to represent measured reality rather than plausible-looking detail.
What Happens Next?
The technique remains an experimental demonstration.
Although the sensor itself does not need to be redesigned, the research uses a custom optical arrangement and calibration process. It is not simply a software update for existing consumer cameras.
Researchers will need to establish how reliably the method works across different sensors, lighting conditions and practical imaging environments.
The important finding is that the physical dimensions of a camera pixel may not tell the entire story about how much spatial information its measurements can reveal.
Original research: Jinchao Liang and colleagues, “Pixel-segmenting imaging physically exceeds pixel-level sampling constraints via intra-pixel quantum-efficiency measurement enabled by evolutionary algorithms,” Communications Engineering, October 15, 2026. Read the peer-reviewed research.
Ray Jackson holds a BSc in Electrical Engineering from the University of Manitoba and a PhD in Physics from Carleton University. His reporting interests include Current and Future Technologies, Engineering and Artificial Intelligence.