Wavelet-based logistic discriminator of dermoscopy images

作者：

Highlights：

• Melanoma dermoscopy features should be based on pixel energies of wavelet filters.

• Selected wavelet bases keep high classification efficiency in low resolutions.

• Logistic learning of dermoscopy images performs well in discriminating melanoma.

• Logistic learning can extract good melanoma classification features.

摘要

•Melanoma dermoscopy features should be based on pixel energies of wavelet filters.•Selected wavelet bases keep high classification efficiency in low resolutions.•Logistic learning of dermoscopy images performs well in discriminating melanoma.•Logistic learning can extract good melanoma classification features.

论文关键词：Feature extraction,Logistic classifier,Wavelets,CAD,Melanoma

论文评审过程：Received 18 April 2019, Revised 15 May 2020, Accepted 12 July 2020, Available online 16 August 2020, Version of Record 10 February 2021.

论文官网地址：https://doi.org/10.1016/j.eswa.2020.113760