CFNN for Identifying Poisonous Plants
نویسندگان
چکیده
Identification of poisonous plants is a hard challenge for researchers because the great similarity between and non- plants. Traditional methods to identify plant can be tiresome, therefore, automated identification system needed. In this work, cascade forward neural network framework proposed based on their leaves. The was evaluated both (poisonous leaves/non-poisonous leaves) which are collected using smart phone internet. Combination shape features statistical extracted from leaf then fed cascade-forward used TRAINLM function training. 500 samples images used, 250 poisonous, remaining non-poisonous.300 in training, 200 testing. Our achieved an accuracy value 99.5%.
منابع مشابه
Poisonous plants.
South Africa is blessed with one of the richest floras in the world, which--not surprisingly--includes many poisonous plants. Theiler in the founding years believed that plants could be involved in the aetiologies of many of the then unexplained conditions of stock, such as gousiekte and geeldikkop. His subsequent investigations of plant poisonings largely laid the foundation for the future Sec...
متن کاملPoisonous native range plants
Plant poisoning is a common problem throughout North America, causing significant economic losses not only through death loss, but through the costs associated with reproductive loss, poor growth rates, herbicides, etc. Through good pasture management, recognizing toxic plants, and by understanding the effects of toxins on animals, plant poisoning can be largely avoided. Plants contain a variet...
متن کاملLiving With Poisonous Plants
The word POISON too often creates fear when it should suggest a warning. Many chemicals or common household substances used incorrectly, usually in the wrong quantity, can cause illness or even death; this is also true with certain plants. Left alone they are harmless. Bruised, crushed, or eaten in varying quantities, they may provoke effects which are upsetting, painful, or even fatal to man.
متن کاملذخیره در منابع من
با ذخیره ی این منبع در منابع من، دسترسی به آن را برای استفاده های بعدی آسان تر کنید
ژورنال
عنوان ژورنال: Baghdad Science Journal
سال: 2023
ISSN: ['2078-8665', '2411-7986']
DOI: https://doi.org/10.21123/bsj.2023.7874