Intelligent recognition method of colour multidimensional image boundary feature based on improved neural network. (2nd September 2019)

• Record Type: Journal Article
• Authors: Ding, Jie; Zhao, Guotao; Hao, Huibing; Li, Chunping
• Is Part Of: International journal of innovative computing and applications. Volume 10:Number 2(2019)
• Page Start: 115
• Page End: 121
• Publication Date: 2019-09-02
• Subjects: improved neural network -- image -- segmentation -- detection -- edge feature extraction; Evolutionary computation -- Periodicals; Neural networks (Computer science) -- Periodicals; Genetic programming (Computer science) -- Periodicals; Biologically-inspired computing -- Periodicals; Swarm intelligence -- Periodicals; Quantum computers -- Periodicals; 006.3
• Languages: English
• ISSNs: 1751-648X

Abstract

In order to improve the intelligent edge feature recognition ability of super-pixel colour multidimensional images, an improved neural network-based edge feature recognition algorithm for colour multidimensional images is proposed. A colour multi-dimensional image imaging model based on super-pixel fusion is constructed, and the edge contour of the multi-dimensional colour image is detected by using the colour block region fusion and segmentation method. The improved neural network adaptive optimisation method is used to segment and match the blocks of multi-dimensional colour images, and to extract the colour boundary information features of the multi-dimensional colour images. According to the RGB value of colour multi-dimensional image and neighbourhood mean, the adaptive fusion segmentation of colour multi-dimensional image is realised. The simulation results show that this method can effectively realise the intelligent recognition of super-pixel colour multi-dimensional image information boundary, the recognition accuracy is higher, and the ability of image recognition and target detection is improved.