Page 547 - e-Book

machine vision by using descriptors of textures and contours, and the usage of algorithms for

deep learning would allow plant species recognition from RGB images especially in the
problem of weed detection in agricultural fields. Due to the nature of problem at hand, the

weed detection has ultimately been defined as the segmentation of the image to weed, plant
and land pixels and state of the art deep learning algorithms have been used for segmentation.

Advanced Image Processing is the use of Computer Algorithms to Perform Image Processing
on Digital Images and “medicinal plant” include various types of plant life used in herbal

medicine[12] its support digital image processing for quality purpose .

Image Pre processing Segmentation Feature
Acquisition from extraction
source

Original data base Image data

Data comparison

Disease detection

Result display

Fig 3.2 Key phases of Digital of image Processing

Fig 3.2 Computerized Image Processing has got parcel of focal points when contrasted
with Analog Image Processing. Picture Processing includes principally four phases to be

specific Image Acquisition, Pre-Processing, Feature Abstraction and Organization Color

Segmentation is the method used to separate crop from the background. Generally weed is a
plant considered undesirable. The ideology is incidentally used to comprehensively portray

species outside the plant realm that can live in different situations and imitate rapidly. Plants
may be generally described in terms of their geometrical, optical and mechanical properties.

Weed classification is a serious issue in Agricultural research. In the olden days weeds in the
agricultural fields were removed with the help of Pesticides. But due to usage of pesticides

537