@article{2759,
  author = {Mohd Norhisham Razali, Noridayu Manshor, Norwati Mustapha, Razali Yaakob, Alfian Abdul Halin, Mohammad Noorazlan Shah Zainudin},
  title = {Improving Invisible Food Texture Detection by using Adaptive Extremal Region Detector in Food Recognition},
  journal = {Journal of Electronic Systems},
  year = {2019},
  volume = {9},
  number = {2},
  doi = {https://doi.org/10.6025/jes/2019/9/2/55-67},
  url = {http://www.dline.info/jes/fulltext/v9n2/jesv9n2_3.pdf},
  abstract = {The advancement of mobile technology with reasonable cost has indulge the mobile phone users to photograph and share their foods in social media. Since that, food recognition has become emerging research area in image processing and machine learning. Food recognition provides an automatic identification of the types of foods from an image. Further analysis of food images can be carried out to approximate the calories and nutrients that can be used for health-care purposes as well as the other application domains. The interest region-based detector by using Maximally Stable Extremal Region (MSER) provides distinctive interest points by representing the arbitrary shape of foods in parallelogram especially the food images with strong mixture of ingredients. However, the classification performance on food categories with less diverse texture food images by using MSER are obviously was not up to par with the rest of food categories that have more noticeable texture.
The respective food objects were suffered from low volume extremal regions (ER) detection that were associated with the condition of food images that have visually texture-less objects, low contrast and brightness as well as small image pixel dimensions. Therefore, this paper proposed an adaptive interest regions detection by using MSER (aMSER) that provide an automatic MSER parameter configuration to increase the density of interest points on the targeted food images. The features
are described by using Speeded-up Robust Feature Transform (SURF) and encoded by using Bag of Features (BoF) model. The classification is performed by using Linear Support Vector Machine and yield 84.20% classification rate on UEC100-Food dataset with competitive volume of ER and extraction time efficiency.},
}