Human Appearances With Medical Properties

Ponnam Mahesh, kodanda Ramarao

Abstract


To be able to deal with the highly dimensional pose space, scene complexity, as well as other human appearances, nearly all existing works require computationally complex training and template matching processes. The positioning, dimensions, and color of the face area can be used for the localization of the body, construction of the models for that lower and upper body based on anthropometric constraints, and estimation of your skin color. We advise a bottom-up methodology for automatic extraction of human physiques from single images, within the situation of just about upright poses in cluttered environments. Segmentation of human physiques in images is really a challenging task that may facilitate numerous applications, like scene understanding and activity recognition. Different amounts of segmentation granularity are combined to extract the pose with greatest potential. Qualitative and quantitative experimental results show our methodology outperforms condition-of-the-art interactive and hybrid top-lower/bottom-up approaches. The segments owed to the body arise with the joint estimation from the foreground and background throughout the part of the body search phases, which alleviates the requirement for exact shape matching.


Full Text:

PDF




Copyright (c) 2017 Edupedia Publications Pvt Ltd

Creative Commons License
This work is licensed under a Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International License.

 

All published Articles are Open Access at  https://journals.pen2print.org/index.php/ijr/ 


Paper submission: editor@eduindex.org