研究目的
Investigating the Adaptive Particle Representation (APR) as a content-adaptive representation of fluorescence microscopy images to overcome storage, memory, and processing bottlenecks.
研究成果
The APR meets all four representation criteria set out in the Introduction, providing a simple and efficient content-aware representation of fluorescence microscopy images. It overcomes storage, memory, and processing bottlenecks, enabling orders of magnitude benefits across a range of image processing tasks.
研究不足
The APR is sub-optimal with respect to the number of particles used, resulting from the conservative restrictions required to derive the efficient Pulling Scheme, and from the generality of the Reconstruction Condition.
1：Experimental Design and Method Selection:
The APR was designed to adaptively resample an image, guided by local information content, using a set of particles with associated intensity values. The methodology includes the computation of an Implied Resolution Function and the use of a Pulling Scheme algorithm to efficiently find optimal solutions to the Reconstruction Condition.
2：Sample Selection and Data Sources:
Synthetic benchmark data in 3D and a corpus of 19 exemplar volumetric fluorescence microscopy datasets of different content and imaging modalities were used for validation.
3：List of Experimental Equipment and Materials:
The open-source C++ APR software library LibAPR was used, compiled with gcc 5.4.0 and OpenMP 4.0 shared-memory parallelism on a 10-core Intel Xeon E5-2660 v3 (25 MB cache, 2.60 GHz), 64 GB RAM, running Ubuntu Linux 16.
4：0 and OpenMP 0 shared-memory parallelism on a 10-core Intel Xeon E5-2660 v3 (25 MB cache, 60 GHz), 64 GB RAM, running Ubuntu Linux Experimental Procedures and Operational Workflow:
04.
4. Experimental Procedures and Operational Workflow: The APR was validated using noisy synthetic benchmark data in 3D, with the performance assessed for synthetic images with numbers of objects roughly corresponding to high, medium, and low complexity images.
5：Data Analysis Methods:
The performance of the APR was assessed in terms of computational and memory costs, with the Computational Ratio (CR) defined as the number of input pixels divided by the number of output particles.
独家科研数据包，助您复现前沿成果，加速创新突破
获取完整内容
