Circular Object Detection Using a Modified Hough Transform

Abstract

A practical modification of the Hough transform is proposed that improves the detection of low-contrast circular objects. The original circular Hough transform and its numerous modifications are discussed and compared in order to improve both the efficiency and computational complexity of the algorithm. Medical images are selected to verify the algorithm. In particular, the algorithm is applied to localize cell nuclei of cytological smears visualized using a phase contrast microscope.

A Model Vision of Sorting System Application Using Robotic Manipulator

“…For object detection, Hough transform is a method widely used for object recognition applications [8]- [10]. Roughly, it checks the likeliness between the cluster and the object model to find what cluster corresponds to that object in an image.…”

A New Approach to Spindle Radial Error Evaluation Using a Machine Vision System

“…The major advantage of this transform is its robustness towards irregularities in detected objects and disturbances like noise under varying illumination [12]. In the proposed method, the contribution of the form error of master cylinder is not taken into account, assuming the shape of the master cylinder to be a circle, for improving the accuracy of spindle radial error evaluation.…”

Segmentation-free quantification of spots on a homogeneous background

“…It illustrates the intended use of our method. In contrast to approaches that have focused on segmenting individual spots [2,3,4,5], our primary goal is to provide fast, but robust global summary measures that quickly allow the analyst to find trends and outliers in image sets from high-throughput microscopy. This should allow them to identify smaller subsets of images of particular interest, to which more sophisticated segmentation-based approaches or manual segmentation can subsequently be applied.…”

“…Analysis of such images requires counting and measuring the sizes of those spots. Even though methods based on wavelet multiscale product (WMP) operator [2], ellipse fitting [3], modified circular hough transform (CHT) [4], or spectral graph partitioning [5] have been introduced to detect and segment individual spots, practitioners still frequently resort to manual analysis because automated segmentation is not reliable enough in many real-world scenarios. In this work, we introduce a novel approach to this problem that derives per-pixel estimates of spot size and density that do not require reliable delineation of individual spots.…”