A NOVEL APPROACH FOR IDENTIFYNG DIFFERENT SHAPES BASED ON EDGES USING CLASSIFICATION ALGORITHMS IN IMAGE PROCESSING

is one of the basic shape analysis and an important task in the area of computer Vision. Shape recognition has main role in supporting many industrial needs, image processing tasks, classification, etc. The work in this project concentrates on shape recognition which specifically starts from image acquisition, edge detection and classification in image processing. The main emphasis is given for basic shape reco and classification with the help of the image processing and classification techniques.


INTRODUCTION
The acquisition and analysis of the visual information produced by the interaction between light and the world objects have represented powerful means through which humans and animals can quickly and efficiently learn about their surrounding environments. In this context, [1] there are several objects, such as characters, which are defined in terms of 2D shapes and should therefore be represented, characterized and processed.
The acquisition and analysis of the visual information produced by the interaction between light and the world objects have represented powerful means through which efficiently learn about their surrounding environments. In this context, [1] there are several objects, such as characters, which are defined in terms of 2D shapes and should therefore be represented,

Computational shape analysis
Computational shape analysis involves several important tasks, from image acquisition to shape classification. [1] Shape acquisition involves acquiring an image (e.g., photograph) and digitizing it, so that it can be properly processed by a computer (Figure 1.3). [4] The type of image acquisition framework and devices to be used depends heavily on the application, e.g., a camera attached to

Shape Pre-Processing
The first step toward the computational morphological analysis of a given object.

Edge Detection
A subsequent important step in most image analysis approaches is the partitioning of the image in terms of its components, or patches that define the different foreground objects and the background of a scene. In order to perform shape analysis over the image objects (i.e., the leaves), it is necessary

Gray-Level Edge Detection Classification
Classification is done particularly in the case of shape analysis as it is important to carefully visualize and inspect the shapes to be classified.

Shape Classification
Finally, after performing shape preprocessing, and Edge detection classification algorithms are usually applied in order to assign a class to each considered shape. The important aspects related to shape classification is Feature Extraction and Image Thresholding. [9]

Steps involved in classification
The stages in shape classification are explained in the following steps.

Figure 6
The concept of edge detection through image differentiation.

Feature Extraction
If the features extracted are carefully chosen it is expected that the features set will extract the relevant information from the input data in order to perform the desired task using this reduced representation instead of the full size input.
Image Thresholding [1]. For instance, in the image, the pixels of the background range from 250 to 255, while those belonging to the objects range from 0 to 240. This is summarized by the following algorithm:

Algorithm: Thresholding
For each pixel in the image do If (pixel value<threshold) Then Label the pixel as an object pixel; Else Label the pixel as a background pixel;

Specification Problem Statement
Existing System: The main features involved are shape preprocessing, edge detection and classification.
There are many systems which are developed using the combination of image preprocessing and edge detection.
There are many systems which are developed using the combination of image preprocessing and edge detection. In this existing system it is identified all types of regular shapes like rectangle, circle, diamond, square, eclipse, etc. and some irreregular shapes like tigers, balls ,apple and combination and classification of these.
Proposed System: In this proposed system, project consists of 3 phases. The 1 st phase is a pre process an image. The 2 nd phase is a edge detection on basic shapes of an input image using 2D transformation techniques. The 3 rd phase is the extracting the features of the images into its classes.

Requirements Specification Purpose of the system:
In this proposed system, project consists of 3 phases. The 1 st phase is a preprocess an image. The 2 nd phase is a edge detection on basic shapes of an input image using 2D transformation techniques. The 3 rd phase is the extracting the features of the images into its classes.

The 3 phases will be carried out as follows: Phase1 Image preprocessing
The query image is converted from RGB to Gray-Scale image by using the MATLAB command rgb2gray. Phase2 Edge detection:

Figure 8
Examples of each of the six classes.
One of the first steps generally required in shape analysis is to detect the shape. In order to perform shape analysis over the image objects, it is necessary to partition image objects and the background of a scene.

Phase3 Shape identification and classification
After performing edge detection on objects in image, calculate the centroid and area of each object in image by using the functions s = regionprops (L, 'centroid') and DT = regionprops (L, 'area'); in MATLAB.The boundaries of an objects in image is found by using the function boundaries = bwboundaries (BW);Now, we find the distances between the boundaries and centroids by applying the function distance{k}(1,i) = sqrt ( ( b(i,2) -s(k).Centroid(1) )^2 + (b(i,1) -s(k).Centroid(2) )^2 ); then find the maximum and minimum distances by using the functions a=max(distance{k}); and b=min(distance{k}); After finding the minimum and maximum distances, compute the difference of an each object in image by using the function difference=a-b; Edge detection, 'feature extraxtion', intensity conversions based upon the input, say rgb2gray or mat2gray etc depending upon the file format ; GIF or JPEG or PNG etc.If threshold of any recognition image matches with the training work space, recognition as that of that Training Image recognized.

Implementation
This project is implemented in three phases: Phase 1: Image Preprocessing Phase 2: Edge Detection

Phase1: Shape Preprocessing
This process involves acquiring and storing an objects in image.Shape detection is performed on image analysis.
1. Here eliminate the noise occurred of an objects in image by using the shape operations. 2. Load the images in the training data set. 3. The objects in the image are retrieved will be given as training images for phase 2, edge detection and phase 3 shape identification and classification.

Phase2: Edge Detection
1. Identify the sets of points (or curves or contours) corresponding to the more pronounced gray-level variations. 2. Calculate the difference between their respective gray-levels. 3. Apply the canny edge detection technique to detect the edges of an each object in image. 4. The generating result will be given for phase 3, shape identification and classification.

Phase3: Shape Identification and Classification
Calculate the area and boundaries of the shapes in the image based on the edges obtained after performing phase 2.
 Compute the distance between the images in the training dataset and the input image by using Euclidean distance.  Obtain the ranges based on area and boundaries of each image.
 The shapes are compared, and values obtained in step 2 should satisfy the threshold value.  The shapes in the image which satisfy the above conditions are identified.  Finally the shapes in the input image can be classified by extracting the features and satisfying the above conditions. Unit testing refers to tests that verify the functionality of a specific section of code, usually at the function level.

System Testing
System testing of software or hardware is testing conducted on a complete, integrated system to evaluate the system's compliance with its specified requirements.

CONCLUSION AND FUTURE WORK
In this work, we identified all the basic shapes and it has been extended to tiger, parrot, lion, mango, strawberry, and tomato images. Even though the project works well for shapes, it has to be extended for irregular shapes and for other domains also .Even though the methodology used works well for only certain shapes like tiger, parrot, lion, mango, strawberry and tomato, it has to be extended to all other irregular shapes by using enhancing the methodology by using WAVELETS, PCA and other techniques for further improvement and classification.