JPH05191762A - Picture information retrieval device - Google Patents

Abstract

PURPOSE:To reduce the retrieval time by displaying a picture on a CRT in an optional size and shape to be identified so as to facilitate the retrieval. CONSTITUTION:Picture information in an HDD 2 is transferred to a picture information expansion section 9 via a HOST use I/F 5, in which the compressed information is expanded and the result is stored in a frame memory section 11. The picture information stored in the memory section 11 is read synchronously with a vertical synchronizing signal and a horizontal synchronizing signal or the like generated by a signal generating section 10 and a color signal is modulated by an encoder 12 and converted into an analog signal by a D/A converter 14, Y/C signals are outputted onto a CRT 16, on which the video image is displayed. When one pattern of the CRT is divided into multi-patterns and a picture is displayed by lots of patterns at the time of retrieval, a control means 1 detects a frequency component of each picture to discriminate the content of each picture and decides a display area (size and shape) on the CRT 16 in a range by which a person identifies the individual picture and operates a CRT controller section 8.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an image information retrieval device for retrieving a desired still image from a storage device storing a large number of compressed still images.

[0002]

2. Description of the Related Art In an electronic still camera, image information relating to a still image picked up by an image pickup device is stored in a removable removable memory card in the form of a digital signal. Then, the memory card is removed from the electronic still camera and attached to the reproducing device, and the image information stored in the memory card is read by the reproducing device. Then, the still image is reproduced and displayed on the television image or the like based on the image information.

The image information of the memory card is transferred to a large-capacity information storage device such as a hard disk, and this storage device constitutes an electronic album (image data file) for the electronic still camera.

When searching for a desired image from a large number of still images stored in this electronic album, there are a method of searching for a still image one frame at a time by a high-speed search, a method of searching using a title as an index, and the like. Also, a system has been proposed in which one screen is divided into a large number of drawings to display a large number of images and a desired image is searched.

[0005]

In the above retrieval system, a system in which a large number of images are displayed on one screen is effective because high-speed retrieval is possible, but there is a limit in distinguishing the displayed images due to the characteristics of the images. Therefore, if the number of divided screens in one screen is increased, the image on each screen becomes smaller than the distinguishable limit, which is a problem.

For this reason, the images must be displayed one after another in a large size, and in the end, it takes a long time to retrieve the images.

An object of the present invention is to provide an image information retrieval device which is easy to retrieve and can shorten the retrieval time.

[0008]

To achieve the above object, the present invention provides an image information storage means for storing compressed image information relating to a large number of images, and an image stored in the image information storage means. In an image information retrieving apparatus including image information decompressing means for decompressing information and display means for displaying an image based on the image information, a frequency component detecting means for detecting a frequency component of each image and this frequency component detecting means A display area determining unit that determines a display area of the image to be displayed on the display unit based on the detected frequency component, and an image is displayed on the display unit in an arbitrary size and shape by receiving an output from the display area determining unit. Image reproduction control means for controlling the reproduction.

The frequency component detecting means is configured to detect the frequency component of each image based on the code amount of the image information, and the shape of the image displayed on the display means is set from the outside. It is characterized in that it is provided with an input means that can be used.

[0010]

According to the above-mentioned means, a large number of images can be displayed on one screen of the display means in a size and shape capable of image identification in accordance with the characteristics of the image to be searched. Images are more likely to be identified and selected on the first screen display, improving search efficiency.

Further, in general, an image having many high frequency components has a fine pattern and the generated code amount at the time of data processing becomes large. Therefore, the frequency component of each image can be easily and reliably detected by detecting the size of the code amount. Can be detected.

Further, by setting the shape of the display image, only the part (for example, the central part) necessary for the image search can be displayed, and a larger number of screens (images) can be displayed on one screen for the search. Efficiency is improved.

[0013]

Embodiments of the present invention will be described below with reference to the drawings.

FIG. 1 is a block diagram showing the overall construction of an embodiment of the present invention. Reference numeral 1 is a host computer (hereinafter abbreviated as HOST) such as a personal computer which is a control means, and 2 is a hard disk provided in the HOST 1. The figure is a hard disk driver (HDD) as an image information storage means such as a driver or an optical disk driver. Further, the HOST 1 is provided with an input device 3 such as a mouse and a keyboard.

A reproducing unit 4 is connected to the HOST 1 via an interface (I / F) 5 for the HOST, and further, the image information photographed by the electronic still camera is subjected to a compression process described later. The stored image information of the memory card 6 is input through the memory card I / F 7.

The reproducing unit 4 is a cathode ray tube (CRT).
A controller unit 8, an image information decompression unit 9 for decompressing compressed image information, a signal generation unit 10 for generating various reference signals, a frame memory unit 11 connected to the CRT controller unit 8, and an encoder 12. , 13, a digital / analog (D / A) converter 14 for Y / C output, and an analog / digital (A / D) converter 15 for Y / C input.

Further, in the figure, 16 is a CRT which is a display means for receiving Y / C output through the D / A converter 14, and 17 is a printer which is connected from the reproducing unit 4 through a digital I / F 18. ..

The basic operation of the above embodiment will be described.

When the memory card 6 is attached to the reproducing unit 4, the compressed image information relating to a large number of images stored in the memory card 6 includes the I / F 7 for the memory card and the I / F 5 for the HOST. It is read into the HOST1 via the HDD and recorded in the HDD2. In this way, HOST1
A large amount of image data files will be created inside.

Next, the operation during reproduction will be described.

The image information in the HDD 2 is the HOST I
It is transferred to the image information expansion unit 9 via / F5. The image information compressed by the image information expansion unit 9 is expanded and stored in the frame memory unit 11. This frame memory unit 11
The image information stored in is read out in synchronism with the vertical and horizontal synchronizing signals generated in the signal generator 10, the color signal is modulated by the encoder 12, and the analog signal is converted by the D / A converter 14. Is converted into the CRT 16 and is output as Y / C to the CRT 16. On the CRT 16, the video is displayed based on the image information.

The digital value relating to the image information is directly output to the printer 17 via the digital I / F 18.

Further, the Y / C input is converted into a digital signal by the A / D converter 15, decoded and stored in the frame memory unit 11, and then input to the image information decompression unit 9 for compression processing. After that, it is stored in the HDD 2 through the I / F 5 for HOST.

The memory card 6 stores the above-mentioned compressed image information. The compression processing of this image information will be described with reference to FIGS.

FIG. 3 is a block diagram showing a structure for compression processing, and 21 is a lens (not shown) of an electronic still camera.
Is a discrete cosine transformer for applying a two-dimensional discrete cosine transform to the captured image information G, 22 is a quantizer for connecting the discrete cosine transformer 21 and linearly quantizing its output D ₁ , and 23 is a quantizer 22. A Huffman encoder connected to the Huffman encoder for encoding the output D ₂ thereof; 24 is connected to the Huffman encoder 23; a code table for supplying the Huffman encoder 23 with parameters required for encoding; 25 is a scaling factor It is a quantization matrix connected to the quantizer 22 via the multiplier 26 to which F is input.

FIGS. 4A to 4C are lenses of an electronic still camera.
FIG. 7 shows a state in which the image information captured in (not shown) is divided into blocks before the discrete cosine transform and a data structure in blocks.

5A to 5C show image information G in block units.
₂ shows the frequency distribution of the information G ₃ when the discrete cosine transform is performed. 5 (a) is image information G ₁ captured by a lens (not shown) of an electronic still camera, FIG. 5 (b) is a block unit image of this image information, and FIG. 5 (c) is this block unit image. It is image information obtained by performing a discrete cosine transform on the information.

The image information G ₁ composed of 768 × 480 pixels shown in FIG. 5A is divided into image information G ₂ of a block unit composed of 8 × 8 pixels, for example, and the electronic information shown in FIG. The image information G captured by the still camera is input to the discrete cosine transformer 21 and the two-dimensional discrete cosine transform is executed.
Discrete cosine coefficients are obtained. Next, the discrete cosine-transformed output D ₁ has a scaling factor F with a different quantization step size for each discrete cosine coefficient.
Using the quantization matrix 25 set via the multiplier 26, the linear quantization is performed by the quantizer 22 to obtain the quantized coefficient.

The quantized coefficient D ₂ output from the quantizer 22 is Huffman-encoded by the Huffman encoder 23 with reference to the code table 24 to generate encoded data D _out .

Encoded data D which is compressed image information
_Out is packaged in blocks as shown in FIG. 4 (c). The data structure of the packaged encoded data _Dout is configured as shown in FIG. 4 (c). 4 (b) shows that one screen (768 × 480 pixels) shown in FIG. 4 (a) is separated into block units B of 8 × 8 pixels, discrete cosine transform is executed for each unit, and then the DC component of each unit is executed. A direct current component (DC) for 96 × 60 blocks is expressed in a conceptual diagram in which These DC components (DC) are arranged at the beginning of each block data forming the image data shown in FIG. 4 (c). At the end of each block, the block end code EOB is arranged in the data area shown in FIG. 4 (c). An image end code EOI indicating the end of all blocks is arranged in the data end area.

The direct current component (DC) of the block data shown in FIG. 4 (c) is arranged at the top of the package, and the alternating current components (AC1, ...) Of the block data are arranged in the data area.
They are arranged in the order of so-called zigzag scanning. AC1 in the data area is 1 of the AC component of the information G ₃ shown in FIG. 5 (c).
That is, the component corresponding to (1, 0) when expressed in frequency coordinates. Next to the EOB in the data area, the DC component (DC) of the next block is arranged, and subsequently 60 × 96 which is all blocks composing one screen are arranged and filed as one screen data. ..

A large number of screen files are stored in the memory card 6 in the above format. Such screen information stored in the memory card 6 is stored in the HDD 2 via the memory card I / F 7 and the HOST I / F 5 in FIG. 1 to form an image data file.

[0033]

[Outer 1]

Next, a search for a multi-screen (image) display is executed from the image information stored in the HDD 2, and a desired screen is selected from the displayed multi-screens, and the image of the desired screen is selected. The operation for enlarging and displaying only one will be described.

FIG. 2 is a block diagram showing a configuration of a main part for executing the above-mentioned multi-screen display and image display of a desired screen.

In the HOST1, the AC component (AC1, ...) Following the DC component (DC) of the block data shown in FIG. 4 (c).
Frequency component detector 3 for detecting the frequency component from
0 and C when receiving the data from the frequency component detector 30
A display area determination unit 31 that determines the size and shape (area) of one of the multiple screens displayed on one screen of the RT16, and a screen selection unit 32 that selects a desired screen from among the multiple screens displayed on the CRT 16. It has and.

FIG. 6 is a flow chart of the multi-screen display. When the image information search is started, the screen files are sequentially read from the HDD 2 (S1-1), and the CRT is performed as follows.
16 Determines the number of screens displayed on one screen.

That is, generally, an image having many high-frequency components has many fine patterns, and an image having many high-frequency components produces a large amount of encoded data D _out . Using this relationship, the HOST 1 detects the frequency component of each image from the frequency of the AC component of the encoded data D _out or by the frequency component detection unit 30 based on the code generation amount, and determines the size T of each screen file. Ask for. This file size T
Is compared with the preset values a _{1 to} a ₄ for determining the size of the screen (image) in the display area determining unit 31 (S1-2),
For example, if a ₁ <T ≦ a ₂
16 is displayed, and if a ₂ <T ≦ a ₃ , 1/8 is displayed and a
_{If 3} <T ≦ a ₄ , 1/4 display is performed, and if a ₄ <T, 1/2 display is determined (S1-3).

The values of a _{1 to} a ₄ are determined based on the identifiable frequency for human images.

When the display area is determined by the display area determining section 31, the CRT controller section 8 of the reproducing unit 4 which is the image reproducing control means displays the CRT 16 with the screen of the determined size.
Control to display an image on (S1-4).

The above operation is repeated for each screen file to display a large number of screens on one screen of the CRT 16.
Then, the number of screens to be displayed becomes the limit when the amount of the file size T of each image is added and the added amount reaches a predetermined fixed amount.

FIGS. 7A to 7D show examples of multi-screen display on the CRT 16. FIG. 7 (a) shows a state where 1/16 display (one screen of CRT16 is divided into 1/16), FIG. 7 (b) shows a state where 1/9 is displayed, and FIG. 7 (c) shows 1/4. Shows the displayed state,
Further, FIG. 7 (d) shows three 1/4 display images and four 1/1 images.
It shows a state where the image of 6 displays and one screen are displayed.

Next, a CR in which multiple screens are displayed as described above
A flowchart relating to the operation for selecting a desired screen from T16 is shown in FIG.

In FIG. 8, when multi-screen display is performed (S2-
1), the screen selection unit 32 outputs a signal to the reproducing unit 4 so as to display a pointing device, for example, a pointer on the screen of the CRT 16 according to an instruction from the input device 3 (S2-2), and selects a desired one from the multiple screens. When the screen is instructed by the pointing device (S2-3 YE
S), the screen selection unit 32 confirms the position of the pointing device.
(S2-4), select the screen file stored in the HDD 2 corresponding to the instructed screen (S2-5), transfer the image information to the reproduction unit 4 (S2-6), The image of the corresponding screen is displayed on the entire CRT 16.

As described above, the number of screens to be displayed on the CRT 16 is determined within a range in which each image can be identified depending on whether or not the image is a detailed display content. Therefore, the desired image is recognized and selected on the first display screen. The probability of being searched is high, and the time required for the search is short.

By the way, there are cases where the image information can be searched for by a part of the displayed image. For example, when a group of images that can be identified even if only the central part of the image is displayed is the search target, it is possible to search by displaying only the face of the person in the center of the image as shown in FIG. 9A. For example, since the amount of information per screen is reduced by displaying only the face, a large number of images (8 × 4) are displayed on one screen of the CRT 16 as shown in FIG. 9B. It is possible to increase the search speed.

FIG. 10 is a flowchart relating to the selection of the display shape of the above screen. When the display shape selection mode is set by the appropriate mode setting means such as the input device 3 (S3-1), the HOST1 To playback unit 4 FIG.
Multiple display shapes with various shapes and sizes as shown in (a)
A signal for displaying 40 on the CRT 16 is output (S3-2). C
When the display form 40 is displayed on the RT16, the input device 3
The pointing device 41 is displayed on the screen of the CRT 16 by an instruction from, and a desired display shape 40 is designated from the plurality of display shapes 40 by this pointing device 41. For example, as shown in FIG. Display form 40 is selected (S3-3).

When the display form 40 is selected as described above, C
Data relating to the display form is output to the RT controller unit 8. The CRT controller unit 8 has a configuration capable of masking an arbitrary area of the image information, and when the image information expanding unit 9 expands the image information from the HOST 1, the masking is performed based on the data related to the display form. Specify the area. Therefore, in the CRT 16, the images on the screen based on the image information in which the display area is limited by the display type 40 are sequentially displayed.

[0049]

As described above, according to the invention described in each claim, the sizes and shapes of a large number of images displayed on the display means are displayed in a state in which the images can be identified. Thus, it is possible to provide an image information search device with improved search efficiency.

[Brief description of drawings]

FIG. 1 is a block diagram showing the overall configuration of an embodiment of an image information search device of the present invention.

FIG. 2 is a block diagram of a main part of FIG.

FIG. 3 is a block diagram showing a configuration for image information compression processing.

FIG. 4 is an explanatory diagram of a block state of image information.

FIG. 5 is an explanatory diagram of a discrete cosine transform of image information.

FIG. 6 is a flowchart relating to multi-screen display.

FIG. 7 is an explanatory diagram showing an example of multi-screen display.

FIG. 8 is a flowchart relating to screen selection.

FIG. 9 is an explanatory diagram of display shape selection.

FIG. 10 is a flowchart relating to display shape selection.

FIG. 11 is an explanatory diagram of a display type and a pointing device.

[Explanation of symbols]

DESCRIPTION OF SYMBOLS 1 ... Control means, 2 ... Image information storage means, 3 ... Input means, 4 ... Image reproduction control means, 6 ... Memory card,
8 ... CRT controller section, 9 ... Image information expansion means,
11 ... Frame memory section, 16 ... Display means, 21 ... Discrete cosine transformer, 22 ... Quantizer, 23 ... Huffman encoder, 30 ... Frequency component detecting means, 31 ... Display area determining section, 32 ... Screen selecting section ..

Claims (3)

[Claims] 1. An image information storage unit for storing compressed image information relating to a large number of images, an image information expansion unit for expanding the image information stored in the image information storage unit, and an image based on the image information. In an image information retrieval apparatus having display means for displaying, a frequency component detecting means for detecting a frequency component of each image and a display of an image to be displayed on the display means by the frequency component detected by the frequency component detecting means. Image information comprising: a display area determining means for determining an area; and an image reproduction control means for receiving an output from the display area determining means and displaying an image on the display means in an arbitrary size and shape. Search device. 2. The image information search device according to claim 1, wherein the frequency component detecting means is configured to detect the frequency component of each image based on the code amount of the image information. 3. The image information retrieval apparatus according to claim 1, further comprising an input unit capable of externally setting the shape of the image displayed on the display unit.