2.1. What Is Information Retrieval Is Important?
2.1 INFORMATION RETRIEVAL
2.1.1 What is Information Retrieval?
The construal of the term information retrieval can be very broad. Just getting a credit card out of your wallet so that you can indite in the card number is a form of information retrieval. However, as an academic field of study, information retrieval might be defined thus: Information retrieval (IR) is finding material (customarily documents) of an unstructured nature (conventionally text) that satiates an information need from within sizably voluminous amassments (conventionally stored on computers).
2.1.2 Why Information Retrieval is needed? Information retrieval is utilized to retrieve germane information resources. Volume of information inclines to become even more sizably …show more content…
The metadata associated with each image can reference the designation of the image, format, color, etc. and can be engendered manually or automatically. This metadata generation process is called audiovisual indexing.
B)Search by example: In this technique, withal called content-predicated image retrieval, the search results are obtained through the comparison between images utilizing computer vision techniques. Amid the test it is inspected the substance of the picture, for example, color, shape, texture or any visual data that can be extricated from the picture. This framework requires a higher computational unpredictability, yet is more productive and dependable than hunt by metadata.
There are picture searchers that amalgamate both inquiry methods, as the primary pursuit is finished by entering a content, and after that, from the pictures acquired can refine the hunt using as hunt parameters the pictures which show up …show more content…
Nevertheless, the outcome of our implementation is not gratifying to any extent. The overall hit-rate is about 43% when HSV model is utilized and about 42% when RGB model is utilized. It is the following reason that makes this approach be inferior to the two verbally expressed above. The first reason the parameters opted for. Since there are two parameters: the number of colors and the distance for evaluation, We may missed the right parameters. Moreover, since the computational intensity is higher than the two above, each change of the parameters will cause a substantial amount of computation. This obviates us to find a better parameter because that will cost me an inordinate amount of time. The second reason is that color correlogram contains two much spatial information, at least much more than CCV. This may be inhibitive in this experiment because the aim of our search is very general. We not only want to retrieve those images of the same objects which may be rotated, or resized; but additionally we optate to retrieve those images of different objects but of the same meaning (i.e. they are all images of buses). Since the requisite is so general, the program may cerebrate two images containing different objects of the same kind (i.e. two images containing a roasted turkey and a roasted lobster, respectively) as images
2.1.1 What is Information Retrieval?
The construal of the term information retrieval can be very broad. Just getting a credit card out of your wallet so that you can indite in the card number is a form of information retrieval. However, as an academic field of study, information retrieval might be defined thus: Information retrieval (IR) is finding material (customarily documents) of an unstructured nature (conventionally text) that satiates an information need from within sizably voluminous amassments (conventionally stored on computers).
2.1.2 Why Information Retrieval is needed? Information retrieval is utilized to retrieve germane information resources. Volume of information inclines to become even more sizably …show more content…
The metadata associated with each image can reference the designation of the image, format, color, etc. and can be engendered manually or automatically. This metadata generation process is called audiovisual indexing.
B)Search by example: In this technique, withal called content-predicated image retrieval, the search results are obtained through the comparison between images utilizing computer vision techniques. Amid the test it is inspected the substance of the picture, for example, color, shape, texture or any visual data that can be extricated from the picture. This framework requires a higher computational unpredictability, yet is more productive and dependable than hunt by metadata.
There are picture searchers that amalgamate both inquiry methods, as the primary pursuit is finished by entering a content, and after that, from the pictures acquired can refine the hunt using as hunt parameters the pictures which show up …show more content…
Nevertheless, the outcome of our implementation is not gratifying to any extent. The overall hit-rate is about 43% when HSV model is utilized and about 42% when RGB model is utilized. It is the following reason that makes this approach be inferior to the two verbally expressed above. The first reason the parameters opted for. Since there are two parameters: the number of colors and the distance for evaluation, We may missed the right parameters. Moreover, since the computational intensity is higher than the two above, each change of the parameters will cause a substantial amount of computation. This obviates us to find a better parameter because that will cost me an inordinate amount of time. The second reason is that color correlogram contains two much spatial information, at least much more than CCV. This may be inhibitive in this experiment because the aim of our search is very general. We not only want to retrieve those images of the same objects which may be rotated, or resized; but additionally we optate to retrieve those images of different objects but of the same meaning (i.e. they are all images of buses). Since the requisite is so general, the program may cerebrate two images containing different objects of the same kind (i.e. two images containing a roasted turkey and a roasted lobster, respectively) as images