Color disease spot image segmentation algorithm based on chaotic particle swarm optimization and FCM
- PDF / 1,602,083 Bytes
- 15 Pages / 439.37 x 666.142 pts Page_size
- 50 Downloads / 208 Views
Color disease spot image segmentation algorithm based on chaotic particle swarm optimization and FCM Lu Xiong1 · Guanrong Tang2 · Yeh‑Cheng Chen3 · Yu‑Xi Hu4 · Ruey‑Shun Chen5
© Springer Science+Business Media, LLC, part of Springer Nature 2020
Abstract Aiming at solving the problems of complex image background and difficulties in the later image segmentation, an image segmentation algorithm based on the chaotic particle swarm algorithm and fuzzy clustering is proposed. First, the color space is converted from the RGB color space into the HIS color space. Then, a hybrid algorithm consisted of chaotic particle swarm optimization and fuzzy clustering is introduced. Each color component is processed by the algorithm, and the corresponding partition graph is obtained. Finally, the color space is converted into the RGB color space to achieve the segmentation effect. Experimental results show that the new algorithm has higher accuracy to segment the image and has good robustness to noises. Keywords Image segmentation · Color disease spot image · Fuzzy C-means · Chaos particle swarm
1 Introduction Since the 1970s, the research of image segmentation has been attracting great attention. Image segmentation has also been the focus and a hotspot in image analysis [1–3]. The extraction of useful information is the foundation of image analysis, and which involves many steps, such as object representation recognition, image segmentation and image preprocessing. In object-type recognition, image segmentation is a key step for achieving successful object recognition. Nowadays, image segmentation is widely used in robotic vision, medical imaging, geographic imaging, object recognition, and other fields, where it has made great achievements [4, 5]. Generally speaking, the algorithms of image segmentation can be classified to four types. Current image segmentation methods can be roughly divided into four categories. The first category is threshold-based image segmentation [6]. Threshold * Ruey‑Shun Chen [email protected] Extended author information available on the last page of the article
13
Vol.:(0123456789)
L. Xiong et al.
segmentation methods are mostly simple and easy to implement. Threshold selection as a key technology directly affects the rationality and effectiveness of segmentation. Because threshold segmentation only considers the gray level of the pixel itself, it pays less attention to the spatial distribution of the pixels and is vulnerable to noise and other factors. The second is image segmentation based on edge detection [7]. The algorithm has fast searching speed and good edge detection, but it cannot get better region structure. There is a contradiction between noise resistance and detection accuracy in edge detection. The third category is image segmentation based on region growth method, and lastly [8, 9], this algorithm can effectively overcome the shortcomings of small and continuous image segmentation space existing in other methods and has better regional features, but it is also prone to over-segm
Data Loading...
