An efficient and robust hybrid method for segmentation of zebrafish objects from bright-field microscope images

Area of Science:

• Computational bio-imaging
• Zebrafish research
• Microscopy image analysis

Background:

• Accurate segmentation of zebrafish in bright-field microscopy is vital for life science applications.
• Partial transparency in early zebrafish stages causes edge ambiguity, challenging traditional segmentation methods.
• Existing computational methods often fail to accurately segment zebrafish edges, leading to measurement errors.

Purpose of the Study:

• To develop an accurate and efficient hybrid method for segmenting zebrafish specimens from bright-field microscope images.
• To overcome the limitations of current segmentation techniques in handling zebrafish transparency and edge ambiguity.
• To provide a robust solution for high-throughput zebrafish image analysis.

Main Methods:

• A hybrid approach combining mean shift algorithm and a distance-regularized level set method.
• Mean shift algorithm enhances image color representation for improved specimen-background discrimination.
• Level set method refines segmentation using initial candidates, followed by heuristic fusion and contour refinement.

Main Results:

• The proposed hybrid method demonstrated improved accuracy and efficiency in zebrafish segmentation.
• Enhanced discrimination between zebrafish specimens and background was achieved.
• The method successfully preserved the explicit contour of the zebrafish object.

Conclusions:

• The hybrid segmentation method offers a significant improvement over existing techniques for zebrafish imaging.
• This approach enhances the reliability of measurements and evaluations in zebrafish studies.
• The developed method is suitable for integration into high-throughput applications in zebrafish research.