Integrating image fusion and motion stabilization for capturing still images in high dynamic range scenes

Kao, Wen-Chung; Hsu, Chien-Chih; Chen, Lien-Yang; Kao, Chia Chuen; Chen, Shou-Hung

doi:10.1109/tce.2006.1706464

Cited by 20 publications

(1 citation statement)

References 12 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Notably, these algorithms heavily rely on extensive data for training the underlying neural network (NN) [11]. Unfortunately, the current commercially available charge-coupled device (CCD) and complementary metal-oxide semiconductor (CMOS) sensors struggle with high dynamic range and produce a vast quantity of dense data making the tracking of high-velocity objects in adverse lighting conditions a challenge [12]- [14]. To address this limitation, event-based cameras, also known as silicon retinas, have emerged as a promising bio-inspired branch of image sensors [15], [16].…”

Section: Introductionmentioning

confidence: 99%

Quantitative Evaluation of a Multi-Modal Camera Setup for Fusing Event Data with RGB Images

Moosmann,

Mandula,

Mayer

et al. 2023

2023 Ieee Sensors

View full text Add to dashboard Cite

Event-based cameras, also called silicon retinas, potentially revolutionize computer vision by detecting and reporting significant changes in intensity asynchronous events, offering extended dynamic range, low latency, and low power consumption, enabling a wide range of applications from autonomous driving to longtime surveillance. As an emerging technology, there is a notable scarcity of publicly available datasets for eventbased systems that also feature frame-based cameras, in order to exploit the benefits of both technologies. This work quantitatively evaluates a multi-modal camera setup for fusing high-resolution dynamic vision sensor (DVS) data with RGB image data by static camera alignment. The proposed setup, which is intended for semi-automatic DVS data labeling, combines two recently released Prophesee EVK4 DVS cameras and one global shutter XIMEA MQ022CG-CM RGB camera. After alignment, stateof-the-art object detection or segmentation networks label the image data by mapping boundary boxes or labeled pixels directly to the aligned events. To facilitate this process, various timebased synchronization methods for DVS data are analyzed, and calibration accuracy, camera alignment, and lens impact are evaluated. Experimental results demonstrate the benefits of the proposed system: the best synchronization method yields an image calibration error of less than 0.90 px and a pixel crosscorrelation deviation of 1.6 px, while a lens with 8 mm focal length enables detection of objects with size 30 cm at a distance of 350 m against homogeneous background.

show abstract