How to: HDR Imaging with Basler Cameras Using Sequencer Mode and PyPylon#

This article describes how to acquire and process HDR (high dynamic range) images using a Basler camera with the Sequencer Mode and the PyPylon SDK. It covers both camera-side configuration and host-side image fusion and tone mapping using OpenCV.

This article is written for ace 2 and boost R cameras, but the principles apply to other Basler cameras with sequencer support as well. Please adapt the code as needed for your specific camera model.

Objective#

The goal is to capture a sequence of images with varying exposure times using the camera's built-in sequencer, and then merge these into a single HDR image with improved detail in shadows and highlights. The result is a tone-mapped image suitable for display or further processing.

Hardware and Software Requirements#

Hardware#

Basler area scan camera (supporting the sequencer mode)
Stable illumination conditions, no flickering lights
Stable scene (fixed camera position and target)

Software#

Python 3.8+
PyPylon Python bindings
OpenCV (cv2)
NumPy

pip install pypylon opencv-python numpy

HDR Capture Workflow#

Exposure Sequence Configuration#

The camera is configured with a list of exposure times in microseconds:

exposure_times_us = [50, 500, 5000, 50000]

Each exposure time is loaded into a separate sequencer set. The camera is configured to cycle through these sets automatically after each trigger.

Key steps:

SequencerConfigurationMode = On
Set ExposureTime and SequencerSetNext
Configure SequencerTriggerSource = ExposureStart
Save each set and exit configuration mode

Once configured, the camera is triggered in Burst Mode:

camera.AcquisitionBurstFrameCount = len(exposure_times_us)
camera.TriggerSource = "Software"
camera.TriggerSelector = "FrameBurstStart"

Captured images are saved as 16-bit TIFFs into a folder, e.g., raw_image/, with file names encoding their exposure time: img_5000us.tiff

HDR Image Processing#

Step-by-Step Process:#

Load 16-bit TIFFs from disk.
Bit-shift to extract 12-bit values.
Normalize to float32 in [0,1].
Apply triangle weight function.
Compute radiance map.
Tone map with Reinhard operator.
Save result as PNG.

Example Images#

The following input images were captured using different exposure times:

Exposure Time	Image
50 µs
500 µs
5000 µs
50000 µs

HDR Fusion Result#

The final tone-mapped image combines the details from all exposures:

Tone-Mapped Image Result

Notes and Recommendations#

Ensure fixed camera position and lighting across all exposures.
Sequencer setup is retained until the camera is powered off or reset.
Use OpenCV EXR support to save true HDR, if required (cv2.imwrite("output.exr", hdr_image)).
For color cameras, additional demosaicing and white balancing steps are needed.

Troubleshooting#

Issue	Cause	Fix
GrabTimeout	Exposure too long or FPS too low	Increase timeout
No HDR effect	Exposure range too narrow	Use more diverse times
File read error	TIFF not written or path wrong	Check permissions and folder

References#

Back to Knowledge Articles