Abstract: Objective Existing steel bar counting methods based on image processing are not accurate and cannotadapt to steel bar diameter variation,or changes in the production process.This paper presents an au-tomatic steel bar counting method which makes use of an iterative Hough round transformation and amulti-threshold connectivity area.The algorithm first extracts ROI by morphological operations.Then,it scales the image size according to the diameter of the input steel bars to the standardizationdiameter size to accommodate the bars diameter change.Nest,an iterative Hough round transforma-tion is applied to the scaled ROI to extract steel bars.Finally,a multi-threshold connectivity areamethod is used to calculate the remaining bars.Experiments show that the algorithm can adapt to thesteel bar diameter changes and changes in the production process with high accuracy.