Hierarchical Queues: general description and implementation in MAMBA image library
نویسندگان
چکیده
Hierarchical queues (HQ) are extremely efficient structures for fast computation of morphological transformations such as watershed lines or geodesic reconstructions. The recent development of MAMBA, new morphological image processing library, led the authors of this paper to review the definition and implementation of the watersheds and geodesic reconstructions algorithms based on these hierarchical queues. This library incorporates very fast implementations of these operators, based on HQ. These implementations increase deeply the performances of MAMBA in real-time applications. Indeed, the documents describing HQ are quite succinct [4, 6]. Unfortunately, these descriptions are far from being accurate and complete. The HQ implementation as it is described in these documents will be used as a starting point for the description of its implementation in the MAMBA library. Regarding the watershed transform, two implementations are available: the first one where only the catchment basins are generated and the second one where these catchment basins are separated by the watershed lines. However, only the first implementation is described in the above mentioned documents. The second one is not addressed although it is more difficult to achieve. The two implementations however are available in some software libraries [7, 8]. The HQ implementation as it is realised in these libraries will be used as a starting point for the description of its implementation in the MAMBA library. Some of its peculiarities and defects will be recalled before indicating which enhancements have been achieved in MAMBA. Regarding geodesic reconstruction, the description of HQ implementations in the above documents exhibits a peculiar feature which makes this algorithm to be tautological. We shall come back to this characteristic and we shall describe an implementation that avoids this problem.
منابع مشابه
Revisiting priority queues for image analysis
Many algorithms in image analysis require a priority queue, a data structure that holds pointers to pixels in the image, and which allows efficiently finding the pixel in the queue with the highest priority. However, very few articles describing such image analysis algorithms specify which implementation of the priority queue was used. Many assessments of priority queues can be found in the lit...
متن کاملA Comparison of Algorithms for Connected Set Openings and Closings
ÐThe implementation of morphological connected set operators for image filtering and pattern recognition is discussed. Two earlier algorithms based on priority queues and hierarchical queues, respectively, are compared to a more recent union-find approach. Unlike the earlier algorithms which process regional extrema in the image sequentially, the union-find method allows simultaneous processing...
متن کاملFuzzy Hierarchical Location-Allocation Models for Congested Systems
There exist various service systems that have hierarchical structure. In hierarchical service networks, facilities at different levels provide different types of services. For example, in health care systems, general centers provide low-level services such as primary health care services, while the specialized hospitals provide high-level services. Because of demand congestion in service networ...
متن کاملFast implementation of large erosions and dilations in Mamba
Large structuring elements are used in many morphological transformations. They are needed when computing size distributions, in filtering operators (and especially in alternate sequential filters), for regularised gradients and large top-hat transforms, for residual operators (skeletons by openings, critical balls, ultimate openings and so on). Moreover, in some applications, images are large ...
متن کاملTowards a general framework for FPGA based image processing using hardware skeletons
In this paper, we present our approach to developing a general framework for FPGA based Image Processing. This framework is based on a library of Hardware Skeletons. A hardware skeleton is a parameterised description of a task-specific architecture. A skeleton’s implementation will apply optimisations specific to the target hardware. The library normally contains a range of alternative skeleton...
متن کاملذخیره در منابع من
با ذخیره ی این منبع در منابع من، دسترسی به آن را برای استفاده های بعدی آسان تر کنید
عنوان ژورنال:
دوره شماره
صفحات -
تاریخ انتشار 2011