INTRDUCTION

The computer vision strategies used to recognize a fruit using basic features
like intensity, color, shape and texture.An important problem in pattern
analysis is the automatic rec- ognition of an object in a scene regardless of its
position, size, and orientation .Zernike moments are the projection of the
image function onto orthogonal basis functions. This paper proposes an
efficient fusion of color and texture features along with zernike moment for
fruit recognition. Color is a useful property which adds the information to
images. The color perceived by human is a combination of three color
stimului such as red (R), green (G), and blue (B), which forms a color space .
However, many color models are used to represent the colors in various
representations such as RGB (red, green, blue), HSV (hue, saturation,
intensity), and CMY (cyan, magenta, yellow) . As compared to monochrome
images, color images have the information of brightness, hue and saturation
for each pixel.
The fruit recognition system has so many applications. The fruit recognition
system can be applied for educational purpose to enhanced learning,
especially for small kids and Down syndrome patients, of fruits pattern
recognition based on the fruit recognition result. It can be used in grocery
store which makes the customers label their purchases using automatic fruit
recognition based on computer vision.it is very difficult to enable a system
for automatic fruit recognition based on images taken from camera.Many kind
of fruits are subject to significant variation in color and texture, depending on
how ripe they are. For example, Bananas range from being uniformly green,
to yellow, to patchy and brown.
The current approaches to invariant two-dimensional shape recognition
include extraction of global image information using regular moments ,
boundary-based analysis via Fourier descriptors or autoregressive models ,
image representation by circular harmonic expansion, and syntactic
approaches. A fundamental element of all these schemes is definition of a set
of features for image representation and data reduction. Normally additional
transformations are needed to achieve the desired invariant properties for
the selected features. After invariant features are computed, they are input to
a designed classification rule to decide a labeling for the underlying image.
The fruits recognition system could be applied as an image contents
descriptor which is able to describe the low level visual features or contents
of the fruit images for the CBIR system. The most popular analys techniques
that have been used for both recognition and classifications of two
dimensional (2D) fruit images are color-based and shape-based analysis
methods. However, different fruit images may have similar or identical color
and shape values. Hence, using color or shape features analysis methods are
still not robust and effective enough to identify and distinguish fruits images
Fruit detection system is primarily developed for robotic fruit harvesting.

However this technology can easily be tailored for other applications such as
on tree yield monitoring, crop health status monitoring, disease detection,
maturity detection and other operations which require vision as a sensor. For
fruit harvesting system, it is very necessary to detect the fruit on the tree
more efficiently. The vision based fruit harvesting system for the fruit
detection basically depend on the contribution of different features in the
image. The four basic features which characterize the fruit are: intensity,
color, edge and orientation. This paper proposes an efficient multiple features
based algorithm for the fruit detection on tree. Color features in image could
be successfully used to segment defects on Jonagold apples are
demonstrated in [8]. Texture features are found to contain useful information
for quality evaluation of fruit and vegetables, e.g., classification of grade of
apples after dehydration with the accuracy of 95% . Color and texture
features are used to locate green and red apples . Combining many features
and classifiers, where all features are concatenated and fed independently to
each classification algorithm