Image analysis – Image transformation or preprocessing – Image storage or retrieval
Reexamination Certificate
2000-07-14
2004-11-02
Couso, Yon J. (Department: 2625)
Image analysis
Image transformation or preprocessing
Image storage or retrieval
C707S793000, C707S793000, C382S282000
Reexamination Certificate
active
06813395
ABSTRACT:
BACKGROUND OF THE INVENTION
This invention relates to an image searching method and an image processing method. More specifically, this invention relates to an image searching method in which specific information extracted from an image such as simple figures or subjects such as human individuals or physical objects or specific information that accompanies the image such as audio information or message information such as characters and marks is stored in a database as accessory information for the image and images are searched through on the basis of the stored accessory information.
The present invention also relates to an image searching method which extracts the subjects of interest on the basis of the position information, camera information and map information that are relevant to the time when the image was taken and which uses the result of extraction (in the form of image) to search through images.
The invention also relates to an image processing method which utilizes such image searching methods or the specific information extracted from the image.
When viewing a portrait, we pay the most attention to the face of the person in the picture. Therefore, when the original image typically recorded on a film is to be printed on a duplicating material such as print paper, the amount of exposure has to be determined such that the face of the person is reproduced in an appropriate color and density.
To meet this requirement of portraiture, Unexamined Published Japanese Patent Application (kokai) No. 346332/1992 proposed a method comprising the steps of dividing an original color image into a multiple of pixels, separating each pixel into three colors R, G and B, performing photometry, constructing histograms about hue and saturation on the basis of the result of photometry, slicing as a human face equivalent area those regions of the histograms which contain pixels having a hue and saturation that correspond to the face of the person in the portrait, and determining the correct exposure on the basis of the photometric data for the sliced regions.
In order to increase the probability for extracting the human face equivalent area, Unexamined Published Japanese Patent Application (kokai) No. 160993/1994 proposed various techniques such as removing the area contiguous to the edges of the image as the background region or transforming the extracted area into a figure defined by lines and determining whether the extracted area corresponds to the face of the person in the portrait on the basis of the shapes of neighboring areas surrounding the extracted area and the shape of the extracted area.
These techniques, however, have a common problem in that if the original image contains the ground, trees and any other areas having a hue and saturation close to those of flesh color, such areas may be erroneously interpreted as the human face equivalent area. If these areas are adjacent the true human face equivalent area, they cannot be correctly separated from the latter, with the result that the original image is not likely to be divided into areas of appropriate sizes.
If the areas other than the face of the person in the portrait are erroneously interpreted as the human face equivalent area or if the areas adjacent the human face equivalent area which have a similar hue and saturation cannot be separated from the latter, it has been impossible to determine the correct exposure for appropriate printing of the true human face.
To solve this problem, it would be effective to utilize not only the color (hue and saturation) of the area of interest but also the information about its shape. Speaking of the human face equivalent area, it will in most cases have an elliptical or oval shape, so this special shape may be added to the color information as a condition for extracting the human face equivalent area.
Thus, the information about the shapes of subjects in the original image is very important to the purpose of image detection or searching. In this connection, reference may be had to the technology disclosed in commonly assigned Unexamined Published Japanese Patent Application (kokai) No. 138471/1997. In this technology, the edges of an image are detected and the lines of contours of an area corresponding to the person in portrait are extracted so that the area, particularly the one corresponding to the face of the person, is rendered to reproduce the preferred color.
According to another prior art image searching system, the image the searcher wants to extract can be searched for without requiring the searcher to draw the image of interest or even if no image close to the one wanted by the searcher is ready at hand or within the list of available images. For this image searching system, reference may be had to Unexamined Published Japanese Patent Application (kokai) Nos. 21198/1995 and 249352/1996.
For use in taking documentary pictures by laymen, cameras have recently been put on the commercial market that provide automatic recording of “when and where” the picture was taken. Those cameras are a composite product in which an artificial satellite based global positioning system (GPS) commonly incorporated in car navigation systems is combined with a direction sensor and a camera [Shashin Kogyo (Photographic Industry), 1996-7, pp. 12-15, 84-87).
While photography has several roles to play, the importance of photo's documentary and evidential performance is one of the essential values of photograph. This is why accessory information telling “when and where” the subject was shot is important to documentary pictures. Conventional cameras have so far provided the information about “when” to some extent but not the information about “when”.
In this sense, the camera described as a new product in Shashin Kogyo, supra (which is hereunder sometimes referred to as a GPS incorporating camera) may well be considered a device that can fulfil the inherent functions of documentary pictures. First of all, the dating function of the conventional “dating” camera and the representation of time by the GPS incorporating camera are entirely different in terms of evidential value: the former provides time that can be determined “arbitrarily” (for personal use) whereas the latter records the “official time” being sent from the satellite.
Exactly speaking, the position information provided by the GPS incorporating camera does not tell the position of the subject but indicates the position of the camera and the direction of shooting. Therefore, if the picture is a long shot, the position of the subject is not always definite. Even in this case and if there is a need, the position of the subject can be correctly calculated from, for example, the distance between that subject and another subject in the image.
The present inventors previously proposed in Japanese Patent Application No. 162130/1998 a photographic printing method comprising the steps of acquiring at least one kind of shooting information in association with the image taken that is selected from among shooting position, shooting direction and magnification rate, obtaining additional information associated with at least a portion of the acquired shooting information and adding the obtained additional information to a photographic print. The present inventors also proposed a photographic printing apparatus implementing the method.
Using the photographic printing method and apparatus, the customer looking at the picture can be easily reminded of relevant atmosphere such as the circumstances of shooting on the basis of the shooting information added in association with the recorded picture. Thus, the method and apparatus can not only enhance the documentary function of photographic prints but also allow them to serve the purpose of amusement (by furnishing a topic of conversation).
SUMMARY OF THE INVENTION
The present invention is a further development of the concept of the above-described prior art for detecting areas of interest using specific shapes in the original image as a key. It is therefore the first object of the invention to provide
LandOfFree
Image searching method and image processing method does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Image searching method and image processing method, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Image searching method and image processing method will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-3281669