Fundamentals of High Dynamic Range (HDR) Imaging

By AZoOptics

Table of Contents

Introduction
Dynamic Range – Definition and Measurement
Characteristics and Image Capture
Dynamic Range with Digital Images
About IDS Imaging Development Systems

Introduction

Although current digital cameras have resolutions more than 10 MP, human sight is comparatively much better than conventional image capturing devices with respect to dynamic range. The human eye is capable of perceiving all brightness levels, but image sensors may lose image data because of overexposure. Nevertheless, High Dynamic Range (HDR) technology enables fine variations in brightness to be imaged even under very bright conditions just as the human eye. This article discusses the fundamentals of HDR imaging technology.

Dynamic Range – Definition and Measurement

HDR imaging represents the capture of digital images with a high dynamic range. On the other hand, images taken utilizing conventional means are referred as Low Dynamic Range or LDR. The dynamic range or contrast of an image is defined as the ration between the largest brightness value and the smallest brightness value. Dynamic ranges are generally represented in the logarithmic unit decibels (dB). The dB value refers the factor by which the highest brightness value is greater than the lowest brightness value. The ratio of the two brightness values (l1 and l2) can be converted into a dB value D with the following equation:

dB values and the corresponding dynamic ranges are listed out in Table 1.

Table 1. Values for dynamic range in dB

Characteristics and Image Capture

The imaging characteristic form plays a crucial role in displaying the variations in brightness when imaging a scene. A system, such as a camera with a conventional CCD sensor, features a linear characteristic if it generates double the output value for twofold the brightness as shown in Figure 1.

Figure 1. Imaging with linear characteristic

Gamma characteristics or gamma curves are power functions of the following form:

They are often utilized in image display on PC screens or in photography. Imaging with gamma characteristic is shown in Figure 2.

Figure 2. Imaging with gamma characteristic

Images with linear characteristic and gamma characteristic are illustrated in Figure 3. An image’s dark areas are brightened by a gamma characteristic as shown in Figure 3. The resulting image matches more to the sensitivity of the human eye. For this, the variations in brightness are condensed in light areas of an image.

Figure 3. Image with linear characteristic (left) and gamma characteristic (right)

Logarithmic characteristic causes even stronger effect and is represented by the following function:

Imaging with logarithmic characteristic is illustrated in Figure 4. Here even greater jumps in brightness in a scene’s light areas cause only minute changes in image brightness. Hence, image sensors with logarithmic characteristics are suitable for capturing images of scenes with very high dynamic range.

Figure 4. Imaging with logarithmic characteristic

Linear characteristics are typically required for image processing such as character recognition and edge detection. On the other hand, the human eye distinguishes variations in brightness on the basis of logarithmic characteristics, which habitually approximate a gamma characteristic in practice.

Dynamic Range with Digital Images

During image digitization, it is necessary to ensure the correct display of the dynamic range of the captured image in the selected image format. The bit depth, based on which the images were digitized, plays a decisive role here. Different bit depths utilizing a gray-scale gradient as an example is shown in Figure 5.

Figure 5. Various bit depths using a gray-scale gradient as an example

A bit depth beyond 8 may be required if further image processing of the digital image data takes place. Bit depth and corresponding possible brightness levels are listed out in Table 2.

Table 2. Bit depth and corresponding possible brightness levels

Contrast adjustment is crucial for HDR sensors, because dull images will be generated due to the fact that the sensor’s dynamic range is often greater than that of the scene. The visual impression of images can be improved with contrast adjustment as shown in Figure 6. The histogram represents the brightness distribution of digital images.

Figure 6. HDR image capture and histogram with minimal contrast (left) and with optimum contrast after a contrast adjustment (right)

The bit depth in the output image is the key for contrast adjustment. Contrast adjustment with 8-bit output data and Contrast adjustment with 12-bit output data are shown in Figure 7 and 8.

Figure 7. Contrast adjustment with 8-bit output data

Figure 8. ontrast adjustment with 12-bit output data

About IDS Imaging Development Systems

IDS Imaging Development Systems, a key manufacturer of digital industrial cameras and frame grabbers, was founded as a ‘two-man firm’ by Jürgen Hartmann and Armin Vogt in 1997. Today the company employs more than 120 staff and is internationally represented by offices in the USA, Japan and France and through their network of distributors in almost all European and Asian countries.

IDS Imaging Development Systems’ customers include OEMs, system integrators and manufacturers in the industrial, security, scientific and medical industries. The company’s products are well-known for their consistent high quality, long-term availability and maximum ease of integration.

This information has been sourced, reviewed and adapted from materials provided by IDS Imaging Development Systems.

For more information on this source, please visit IDS Imaging Development Systems.