Abstract

A new approach to computing the gradient characteristics of a grayscale image as an array of features of an object of interest is considered. It is proposed to design a model of a reconfigurable computing environment that can simultaneously process each pixel of the source image in parallel mode and generate an array with gradient characteristics. Due to the architectural principles of model construction, the gradient is computed in a single clock cycle of the elementary calculator of the reconfigurable computing environment.

中文翻译：

---

一种确定图像梯度特征的可重构计算环境模型设计

摘要

考虑了一种将灰度图像的梯度特征计算为感兴趣对象的特征阵列的新方法。提出设计一个可重构计算环境的模型，该模型可以并行方式同时处理源图像的每个像素，并生成具有梯度特征的阵列。由于模型构建的架构原理，在可重构计算环境的基本计算器的单个时钟周期内计算梯度。