[Rupam]

Problem Statement
Interest in the potential of digital image processing for image recognition has increased enormously over the last
few years. Various developments are made towards automated machine learning for object recognition. Where various approaches are made towards object recognition such as food samples, grain samples, foreign object, and medical
applications etc. The automated ecological learning is still remain untouched. The problem with the study of
ecological learning such as birds, animals or aquamarines creature is the variable nature of the sample. The ecological samples were found to have major variation such as color, homogeneity, and shapes from species to species. This
problem get still immense with aquamarine species such as fishes, where each species varies with others in very less difference such as the fin structure, tail structure etc. The extraction of content features with local and global
feature and their identification for various fish samples is proposed in this project work.





OBJECTIVES OF STUDY


The objective of this project work is to develop a automated aquamarine learning system for fish recognition. This project work process images of different types of fish samples, extracting the features of the fish samples and developing a suitable knowledge based classifier model to classify and recognize the different types
of fish images using three different types of feature sets, viz., color, texture and shape, and finally compare their performance. The Content Based Image Retrieval system developed uses visual content to search images from
large-scale image database according to users interest, retrieving similar images from the database, when an image is given as a query.





MOTIVATION FOR THE PRESENT WORK


Object recognition / classification are the most complex tasks to be performed by the system and require the
large computational effort. knowledge based classifier is useful in lowering the requirement in computer resources for object differentiation, if they are successfully trained. Content-based image retrieval is desirable in many
applications. One of the difficult problems is the ease with which human eyes capture the image content are not understood well enough to automate it.


Problem arises in the selection of the feature and its extraction from the image that capture the perceptual and semantic meaning from the images. Other problems arise in efficient matching of features, storage and indexing. These
problems make the aquamarine learning for fish makes a challenging task.





SCOPE OF THE STUDY


The fish type classification using images of samples is attempted here using Color, Shape and Textural features and results of such a system is applicable to similar types of image, which can perform identity check of a fish. It is well know that conventional fish database can only be retrieved by text query. But image retrieval does not exist. This
project work develops an automated learning for aquamarine fish database for the learning and recognition of fish image, which will find it major application in the area of distance learning. The fish image captured at the
seaside can be learned easily through the developed system, which could have been a tedious task under manual learning.


Database:
  images.zip (Size: 1.02 MB / Downloads: 325)
Code:
  code.zip (Size: 10.39 KB / Downloads: 284)
Read Me:
  readme.doc (Size: 26.5 KB / Downloads: 208)
Report:http://grasshoppernetwork.com/showthread.php?tid=342

[harryyao]

Are you sure this matlab codes can work, because I have try it several times and found it cannot work. my Matlab version is (R2001a)

[Rupam]

(08-29-2012 01:09 AM)harryyao Wrote:  Are you sure this matlab codes can work, because I have try it several times and found it cannot work. my Matlab version is (R2001a)

It is tested in Matlab 2006 and does work.

Probably you could put a snapshot of the errors that you see! And be sure that you have the full version with proper toolboxes in place. I feel it is a version problem. However I cant be sure unless I know where you are facing a problem.

[harryyao]

Dear Rupam

thanks to reply. I still a little bit confused about that. First we should do
>> cd Fish

the error is like as follow:

??? Undefined function or variable 'fish'.

that means it should be a Fish.m file in the code file, but actually It is not existed in there.

In the readme.doc it is said that All the training
images are stored in FDB1 folder. but that folder is not existed too.

Could you upload the fold again, many thanks. Regards

[Rupam]

Ok, I will look at the error.
However one quick note: cd fish does not corresponds to any fish.m. It points to the fish directory.

[fatima siddiqui]

hello Rupam i want to notify that there is no fish folder in the code folder which i have downloaded from the link http://grasshoppernetwork.com/showthread...42..please help as i really do wanna try ur codes

(10-29-2011 04:26 PM)Rupam Wrote:  Problem Statement
Interest in the potential of digital image processing for image recognition has increased enormously over the last
few years. Various developments are made towards automated machine learning for object recognition. Where various approaches are made towards object recognition such as food samples, grain samples, foreign object, and medical
applications etc. The automated ecological learning is still remain untouched. The problem with the study of
ecological learning such as birds, animals or aquamarines creature is the variable nature of the sample. The ecological samples were found to have major variation such as color, homogeneity, and shapes from species to species. This
problem get still immense with aquamarine species such as fishes, where each species varies with others in very less difference such as the fin structure, tail structure etc. The extraction of content features with local and global
feature and their identification for various fish samples is proposed in this project work.





OBJECTIVES OF STUDY


The objective of this project work is to develop a automated aquamarine learning system for fish recognition. This project work process images of different types of fish samples, extracting the features of the fish samples and developing a suitable knowledge based classifier model to classify and recognize the different types
of fish images using three different types of feature sets, viz., color, texture and shape, and finally compare their performance. The Content Based Image Retrieval system developed uses visual content to search images from
large-scale image database according to users interest, retrieving similar images from the database, when an image is given as a query.





MOTIVATION FOR THE PRESENT WORK


Object recognition / classification are the most complex tasks to be performed by the system and require the
large computational effort. knowledge based classifier is useful in lowering the requirement in computer resources for object differentiation, if they are successfully trained. Content-based image retrieval is desirable in many
applications. One of the difficult problems is the ease with which human eyes capture the image content are not understood well enough to automate it.


Problem arises in the selection of the feature and its extraction from the image that capture the perceptual and semantic meaning from the images. Other problems arise in efficient matching of features, storage and indexing. These
problems make the aquamarine learning for fish makes a challenging task.





SCOPE OF THE STUDY


The fish type classification using images of samples is attempted here using Color, Shape and Textural features and results of such a system is applicable to similar types of image, which can perform identity check of a fish. It is well know that conventional fish database can only be retrieved by text query. But image retrieval does not exist. This
project work develops an automated learning for aquamarine fish database for the learning and recognition of fish image, which will find it major application in the area of distance learning. The fish image captured at the
seaside can be learned easily through the developed system, which could have been a tedious task under manual learning.


Database:
Code:
Read Me:
Report:http://grasshoppernetwork.com/showthread.php?tid=342

[Rupam]

It's a separate image folder.

[wan.ahmad82]

(04-19-2013 11:43 PM)Rupam Wrote:  It's a separate image folder.

Hi Rupam..thanks for providing the code for us...i tried to run the code in command windows in Matlab2008 but unable to see the graphical user interface when i type 'top'. Can u advise me? thanks

[wan.ahmad82]

(05-26-2013 10:52 PM)wan.ahmad82 Wrote:  Dear sir,

(05-26-2013 10:52 PM)wan.ahmad82 Wrote:  May i know where is FDB1 folder? may i get that folder too

(05-26-2013 10:52 PM)wan.ahmad82 Wrote:  

(04-19-2013 11:43 PM)Rupam Wrote:  

[hankichul]

anybody have sample data /coding fish recognition in matlab ann

---

anybody can help me to give dome sample fish recognition code/data using matlab 7.9 in ann