Deep Self-Organizing Map Neural Networks for Plantar Pressure Image Segmentation Employing Marr-Hildreth Features

Deep Self-Organizing Map Neural Networks for Plantar Pressure Image Segmentation Employing Marr-Hildreth Features

Jianlin Han (Glorious Sun Guangdong School of Fashion, Huizhou University, Huizhou, China & Huidong Shoes Science and Technology Innovation Center, Huizhou, China), Dan Wang (Tianjin Key Laboratory of Process Measurement and Control, School of Electrical Engineering and Automation, Tianjin University, China), *Zairan Li (Wenzhou Polytechnic, Wenzhou, China & Tianjin Key Laboratory of Process Measurement and Control, School of Electrical Engineering and Automation, Tianjin University, China & Glorious Sun Guangdong School of Fashion, Huizhou University, Huizhou, China) and Fuqian Shi (Rutgers Cancer Institute of New Jersey, New Brunswick, USA)
Copyright: © 2021 |Pages: 21
DOI: 10.4018/IJACI.2021100101
OnDemand PDF Download:
No Current Special Offers


Using the plantar pressure imaging analysis method to realize the optimization design of shoe last is still relatively preliminary. The analysis and utilization of imaging data still have problems such as single processing, incomplete information acquisition, and poor processing model robustness. A deep self-organizing map neural network based on Marr-Hildreth filter (dSOM-wh) is developed in this research. The structure and learning algorithms were optimized by learning vector quantization (LVQ) and count propagation (CP). As a kind of Marr-Hildreth filter, Laplacian of Gaussian (LoG) was developed for the preprocessing. The proposed method performed high effectiveness in accuracy (AC) (92.88%), sensitive (SE) (0.8941), and f-measurement (F1) (0.8720) by comparing with ANN, CNN, SegNet, ResNet, and pre-trained inception-v neural networks. The classification-based plantar pressure biomedical functional zoning technologies have potential applications in the comfort shoe production industry.
Article Preview

1. Introduction

The research of plantar pressure has become a hot-issue in biomechanics. The pressure distribution of the human body’s plantar reflects the foot structure and function, and it controls the whole body posture. Testing and analyzing plantar pressure may obtain the human body’s physiological, pathological, and functional parameters under various poses and exercises. It is of great significance for clinical medical diagnosis, disease determination, postoperative efficacy evaluation, biomechanics, and rehabilitation research (Dissaneewate T., Rungsri T. N., Cheunchokasan P., Leelasamran W. 2021; Korada H., Maiya A., Rao S. K., Hande M. 2020). The analysis and research work of plantar pressure can analyze human foot physical force’s change during walking. This process can observe the degree of foot inversion and eversion and the overall foot in walking-movement lines. Excessive foot varus or valgus may cause a certain degree of a foot injury; in this case, the foot’s excessive rotation can also lead to erratic foot movement. Plantar pressure can also accurately diagnose flat feet, arch collapse, walking elasticity, shock absorption, hip joint and waist injuries of the foot varus, etc., which provides a fundamental data-base for shoe design (Cesar G. M., Buster T. W., Burnfield J. M. 2020; Lan K., Fong S., Liu S.L, Wong R.K., Dey N., Millham R.C. & Wong K.K.L. 2021).

The current research has still been investigated at the subjective evaluation stage; analysis methods are linear. The usual approach is to divide the plantar force area into several sub-areas. The previous design, which is crucial to the finished shoe comfort, cannot be practical and in-depth. The image segmentation of plantar pressure directly affects the accurate recognition and understanding of the plantar interest area. In the existing image processing research, various problems in image segmentation and edge detection are gradually being solved and discussed in depth by researchers (Chaki J & Dey N. 2020). Related theoretical methods and experimental research have made considerable progress. A lot of processing methods have been proposed for images in different types and structures; and specific results have been achieved in various applications (Li Z., Wang D., Dey N., Ashour A. S., Sherratt R. S., Shi F. 2019; Wang D., Li Z., Dey N., Ashour A. S., Moraru L., Sherratt R. S., Shi F. 2020). For example, Machado et al. proposed a design guidance manual for better design of insoles and shoes. To achieve therapeutic intervention in advance, Machado (Machado Á. S., Bombach G. D., Duysens J., Carpes F. P. 2016) found that comfortable shoes have not the same feeling in every foot area; while the toe area has a significant impact on the wearing comfort of the shoe, metatarsophalangeal preferred area (MPA), and arched area; furthermore, they obtained the essential footwear for designing comfortable shoes, insoles, and foot touch products. Rodgers explained that in the place where the toes leave in walking, the tension across the longitudinal arch would increase to provide foot stability. The foot’s sensitivity can be reflected by determining the sensitivity of foot pressure and evaluating various foot positions. There is no clear explanation of the practical and accurate method of extracting sensitive areas (Witana C. P., Goonetilleke R. S., Xiong S., Au E. Y. L. 2009; Roscoe D., Roberts A.J., Hulse D., Shaheen A., Hughes M.P., Bennett A. 2018).

Complete Article List

Search this Journal:
Open Access Articles
Volume 13: 6 Issues (2022): Forthcoming, Available for Pre-Order
Volume 12: 4 Issues (2021)
Volume 11: 4 Issues (2020)
Volume 10: 4 Issues (2019)
Volume 9: 4 Issues (2018)
Volume 8: 4 Issues (2017)
Volume 7: 2 Issues (2016)
Volume 6: 2 Issues (2014)
Volume 5: 4 Issues (2013)
Volume 4: 4 Issues (2012)
Volume 3: 4 Issues (2011)
Volume 2: 4 Issues (2010)
Volume 1: 4 Issues (2009)
View Complete Journal Contents Listing