HDR imaging, or High Dynamic Range imaging, has become increasingly important when selecting industrial cameras. While resolution and frame rate were traditionally considered the primary criteria for camera selection, sensitivity and dynamic range are now critical factors—especially in automotive applications.

Dynamic range refers to the ability of a sensor to capture both bright and dark areas within an image without losing details. This is particularly crucial in real-world scenarios with significant brightness variations (such as driving). For instance, when a car drives out of a tunnel into bright daylight, sensors with low dynamic range often produce images that are either underexposed or overexposed. This results in lost detail in both dark and bright areas—which can be problematic for driver-assistance systems relying on such data.

To maximize the acquisition of important details from very bright as well as very dark regions, achieving a high level of dynamic range is essential.

Two Approaches to Increasing Dynamic Range

There are two fundamental approaches to increasing the dynamic range of final images: hardware improvements and software algorithms.

Hardware Improvements for Higher Dynamic Range

The dynamic range of CMOS sensors depends on several factors. One key element is the sensor’s sensitivity or sensor responsivity, which relates to how well a sensor can capture light across its entire operational range. A higher level of sensor responsivity enables better differentiation between bright and dark areas within an image.

CMOS sensors from manufacturers like Sony Pregius have incorporated multi-exposure functions, allowing users to capture multiple native images with different exposure times without manually changing settings between shots—making HDR processing more efficient for end-users.

Software-Based Solutions

Software-based methods involve capturing a series of bracketed exposures at different exposure levels and combining them into an HDR image. These techniques include:

  • Time Varying Exposure (TVE): This technique uses multiple images captured with varying exposure times to generate HDR results.
  • Spatially Varying Exposure: Here, certain groups of pixels are exposed at different times.

Hardware-Based Solutions

To avoid artifacts caused by multi-exposure techniques, modern sensors incorporate spatially varying exposure (SVE) technology. This technique exposes specific pixel groupings on the sensor using different exposures to achieve an image with a wider dynamic range without losing resolution or introducing unwanted artifacts like ghosting from moving objects.

HDR Image Display and Tone Mapping

When displaying HDR images, one common challenge is that display devices have limited dynamic ranges compared to human visual perception. While HDR displays are available in some cases, they aren’t yet widely adopted.

If an HDR image needs to be displayed on a device with lower dynamic range, tone mapping must reduce the image’s dynamic range through a process called tone mapping. This involves reducing the dynamic range while preserving as much information and contrast as possible.

There is no standard procedure for implementing this reduction since it depends primarily on achieving specific goals—whether that means an accurate representation of scene characteristics or subjective artistic quality.

Tone Mapping Techniques

Two main types exist: global tone mapping applies a uniform transformation to all pixels, making these algorithms efficient enough for real-time processing. Local tone mapping focuses locally and can preserve contrast better but requires more computational resources.

The Imaging Source has long supported HDR image data acquisition through its end-user software IC Measure by incorporating native exposure series support (sensor-dependent) and enabling automatic HDR functionality with minimal user input.

This article, written by Dr. Oliver Fleischmann from The Imaging Source, was originally published in the April 2018 edition of the German-language industry journal inspect under the title “High-Dynamic-Range Imaging in modernen Industriekameras.”

Last Updated: 2025-09-05 00:40:39