JP2011223244A - Data compression device and data decompression device - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a data compression device capable of performing more efficient compression without any waste of arithmetic operation than conventional compression technique utilizing features of target image data, and to provide a data decompression device for decompressing the compressed data.SOLUTION: The data compression device, which applies lossless compression processing to image data obtained by photographing a target to be measured in one or more wavelength bands and constituted by a succession of pixel value per pixel, the pixel value being represented by a predetermined bit number N (N: a natural number), comprises: image data dividing means for dividing the pixel values of a single pixel into high-order bit data and low-order bit data at a predetermined bit number for division, the number being smaller than the bit number N; determination means for determining whether at least one of the divided bit data is suitable for compression; and compression means for compressing at least one of the high-order bit data and low-order bit data of image data which is of the same kind of or similar to the aforementioned processed image data if at least one of the divided bit data is suitable for the compression.

Description

  The present invention is image data captured in one or a plurality of types of wavelength bands, and consisting of a series of pixel values for one pixel, which is a numerical value represented by a predetermined N (N is a natural number) bit number. BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a data compression apparatus that performs reversible compression processing and a data decompression apparatus that restores the compressed data, and in particular, a data compression apparatus that performs high-speed and high-compression processing by focusing on the characteristics and characteristics of image data, and the compression The present invention relates to a data restoration device for restoring data.

  2. Description of the Related Art Conventionally, techniques for compressing data such as image data have been widely used to reduce storage capacity and communication volume, and a data compression apparatus using these techniques has been realized.

Here, the purposes of compressing image data are mainly (1) and (2).
(1) The purpose of compressing data and storing a large amount of data in a limited storage (such as a hard disk) (2) The purpose of compressing data and transferring it at a higher speed at a limited line speed. Here, the line means general input / output from the network or storage.

In the case of the purpose (1), it is advantageous to employ an algorithm having an excellent compression rate even if it takes some time, particularly for data backup for long-term storage.
In the case of (2), when data is used immediately after transfer, such as browsing a WWW (World Wide Web) page or analyzing image data, the decompression time is short even if the compression rate is not somewhat good. May be more advantageous.
That is, the data compression technique has three performance indexes, ie, compression rate, compression time, and decompression time, and can be selectively used according to the purpose.

Image compression techniques used by conventional data compression apparatuses include JPEG (Joint Photographic Experts Group), which is an irreversible compression method for still images, and MPEG (Moving Picture Experts Group), which is an irreversible compression method for moving images.
Conventional data compression apparatuses are also widely applied to image data using algorithms such as LZW, ZIP, Deflate, and Bzip2 that are general-purpose data lossless compression algorithms.

  As a prior art document related to such a data compression apparatus and data restoration apparatus, there is Patent Document 1 below.

JP 2006-279207 A

  However, in a conventional data compression apparatus, when image data is compressed using a compression technique (LZW, ZIP, Deflate, Bzip2), it is generally possible to realize an excellent compression ratio, but generally it is necessary for compression and expansion. There was a problem that the time was long.

  For example, in a conventional data compression apparatus, when compression is performed on 16-bit grayscale image data in which one pixel is one color, the background of the image data is caused by readout noise, dark current, or the like caused by the image sensor. There is a problem in that compression is difficult because it has random properties due to noise and noise of the received signal and is not regular.

  Therefore, (1) it is suitable for the purpose of reducing the data amount as much as possible and storing it for a long time, but (2) it is suitable for the purpose of using an image with a short response time by extending in a short time. There was a problem that there were not many cases.

  In addition, these general-purpose algorithms try to compress a large amount of CPU (Central Processing Unit) time (time for the CPU to perform calculation) even for data that is difficult to compress in principle. There was a problem that the efficiency was not good in terms of compression rate.

  Further, the conventional data restoration apparatus has a problem that the decompression speed is slow and the restoration takes time to restore the compressed data of the image data which has poor regularity and is difficult to compress.

  The present invention solves such problems, and an object of the present invention is to use the characteristics of target image data and perform efficient compression with less wasteful computation than conventional compression techniques. An object of the present invention is to realize a data compression device and a data restoration device that restores the compressed data.

In order to achieve the above object, the invention described in claim 1 of the present invention is:
Image data obtained by photographing the measurement object in one or more types of wavelength bands, and reversibly converted to image data composed of a series of pixel values for one pixel, which is a numerical value represented by a predetermined N (N is a natural number) bit number. In a data compression apparatus that performs compression processing,
Image data dividing means for dividing the pixel value for one pixel into upper bit data and lower bit data at a predetermined number of divided bits smaller than the N bits;
Determining means for determining whether at least one of the divided bit data is suitable for compression;
When at least one of the divided bit data is suitable for compression, a compression unit that compresses at least one of the upper bit data or the lower bit data of image data captured in the same wavelength band as the image data; A data compression apparatus comprising:

The invention according to claim 2 is the data compression apparatus according to claim 1,
The determination means includes
The determination means includes
A compression process in which at least one of the following (1) to (5) is performed on pixel values of a plurality of predetermined pixels in the image data, and the calculated compression ratio is the minimum compression ratio or A compression process in which the measured compression time is the shortest time is extracted, and it is determined that at least one of the upper bit data and the lower bit data compressed by the extracted compression process is suitable for compression. And
(1) Compression of only the upper bit data (2) Compression of only the lower bit data (3) Compression of undivided bit data (4) Compression of the upper bit data and the lower bit data (5) Division or not No compression of divided bit data

The invention according to claim 3 is the data compression apparatus according to claim 1 or 2,
When the compression means compresses only the upper bit data or the lower bit data, the compression means comprises storage means for storing the compressed data and the lower bit data or the upper bit data that are uncompressed in association with each other,
The image data dividing means is
Before the pixel value for one pixel is divided into the upper bit data and the lower bit data, subtracting means for subtracting the minimum value of the pixel value in the image data from each pixel value, To do.

The invention according to claim 4 is the data compression apparatus according to claim 2 or 3, wherein
The image data dividing means divides the image data into data arrays corresponding to rows or columns, divides each pixel value in these rows or columns into the upper bit data and the lower bit data,
The determination means performs at least one of the compression processes (1) to (5) for at least one of the upper bit data or the lower bit data for each of a plurality of predetermined rows or columns. A compression process having a compression ratio larger than a predetermined threshold is extracted from the compression processes of the above, and at least one of the upper bit data, the lower bit data, and the undivided bit data compressed by the extracted compression process Is determined to be incompatible with compression,
When it is determined that at least one of the upper bit data or the lower bit data is suitable for compression, the compression means is adapted to compression for each row or column of the same or similar image data as the image data. The determined upper bit data or at least one of the lower bit data is compressed.

The invention according to claim 5 is the data compression apparatus according to claim 2 or 3,
The image data dividing means divides the image data into blocks each composed of a plurality of pixels and smaller than a range occupied by one unit of a measurement target, and sets each pixel value in these blocks Divided into bit data and the lower bit data,
The determination means performs at least one of the compression processes (1) to (5) for at least one of the upper bit data and the lower bit data for each predetermined block,
When it is determined that at least one of the upper bit data and the lower bit data is suitable for compression, the compression unit determines that the same or similar image data as the image data is suitable for compression for each block. Further, at least one of the upper bit data and the lower bit data is compressed.

The invention described in claim 6
Reversible compression processing is performed on image data captured in one or a plurality of wavelength bands and composed of a series of pixel values for one pixel, which is a numerical value represented by a predetermined N (N is a natural number) bit number. In the data decompression device which decompresses the compressed data compressed by the data compression device according to any one of claims 1 to 3,
Stores compressed data of image data compressed by the data compression device, and only one of upper bit data and lower bit data divided by the data compression device at a predetermined division bit number smaller than the N bits Storage means for storing the compressed data and the uncompressed lower bit data or the upper bit data in association with each other,
A data decompression device comprising decompression means for decompressing the compressed data and combining the uncompressed upper bit data or the lower bit data stored in the storage means in association with the compressed data .

The invention according to claim 7 is the data compression device according to any one of claims 1 to 5 or the data restoration device according to claim 6,
The image data is image data obtained by irradiating a cell sample with excitation light, imaging the cell for drug discovery screening by performing image processing based on a fluorescence signal from the cell, in a plurality of types of wavelength bands. It is characterized by that.

  According to the data compression device of the present invention, using the characteristics and properties of image data, the dividing means divides the data into upper bit data and lower bit data, and at least one of the bit data divided by the judging means is suitable for compression. If at least one of the bit data divided by the compression unit is suitable for compression, the image data is the same or similar to the image data (for example, the image data was shot in the same or the same wavelength band). By compressing at least one of the upper bit data or lower bit data (such as image data), there is less waste in computation than conventional compression techniques (the amount of computation to achieve an equivalent compression rate is small and high speed) ), Efficient compression can be performed.

  According to the data restoration device of the present invention, when the data compression unit compresses only one of the upper bit data and the lower bit data divided at a predetermined number of divided bits smaller than N bits, this compression is performed. Storage means for storing data and uncompressed lower bit data or upper bit data in association with each other, decompressed compressed data, and storing uncompressed upper bit data or lower bits stored in association with compressed data in the storage means By providing a restoring means for combining data, there is less waste in computation than the conventional technique (the computation amount for achieving the same compression rate is small and high speed), and efficient restoration can be performed.

  Further, according to the data compression device of the present invention, the dividing means divides the image data into data arrays corresponding to the rows or columns, and the pixel values in these rows or columns are converted into upper bit data and lower bit data. By dividing into two, the compression means can compress the individual one-dimensional arrays in parallel, which is effective in that the speed can be easily increased.

  Further, according to the data compression apparatus of the present invention, the dividing unit divides the image data into a plurality of blocks each of which is a block made up of a plurality of pixels and is smaller than the range occupied by one unit of the measurement object. Each pixel value in the block is divided into upper bit data and lower bit data, and the determination means determines whether or not the image data is suitable for compression for each “block”, and the compression means is the same as the image data. By compressing image data shot in the wavelength band for each “block”, the regularity on the upper bit side is increased, and the compression rate can be improved.

1 is a configuration explanatory view showing an embodiment of a data compression apparatus of the present invention. It is explanatory drawing of the storage method of the bit data divided | segmented by the division means of FIG. FIG. 2 is an operation flow diagram in which the data compression apparatus of FIG. 1 compresses data relating to a large number of images. It is explanatory drawing of the determination operation | movement by the determination means of FIG. It is an operation | movement flowchart at the time of the data compression apparatus of FIG. 1 compressing data. 1 is a configuration explanatory diagram of an embodiment of a data restoration device of the invention. It is an operation | movement flowchart at the time of the data decompression | restoration apparatus of this invention decompress | decompresses (decompresses) compressed data. FIG. 2 is a configuration explanatory view showing an embodiment of an image analysis system using the data compression apparatus of FIG. 1.

<First embodiment>
(Configuration overview)
Hereinafter, the present invention will be described in detail with reference to the drawings. FIG. 1 is an explanatory diagram showing a configuration of an embodiment of a data compression apparatus according to the present invention, and FIG.

  In FIG. 1, a data compression apparatus 1 according to the present invention is a pixel for one pixel, which is image data taken in a plurality of types of wavelength bands and is a numerical value represented by a predetermined N (N is a natural number) bits. A reversible compression process is performed on image data consisting of a series of values.

Hereinafter, in order to simplify the description, it is assumed that the pixel value for one pixel is represented by 16 bits and is a grayscale image of one color.
Note that the data compression apparatus of the present invention can be applied even when one pixel has a pixel value represented by the number of bits other than 16 bits (such as 32 bits) or a color image.

  The data compression apparatus of the present invention mainly includes a dividing means (image data dividing means) 11 for dividing a pixel value for one pixel into upper bit data and lower bit data at a predetermined divided bit smaller than 16 bits. And determining means 12 for determining whether at least one of the divided upper bit data and lower bit data is suitable for compression, and when at least one of the divided bit data is suitable for compression, the image data Compression means 13 for compressing at least one of the upper bit data and the lower bit data of image data photographed in the same wavelength band, and when the compression means 13 compresses only the upper bit data or the lower bit data, this compressed data And storage means 14 for storing uncompressed lower bit data or upper bit data in association with each other , Comprising a.

  The data compression apparatus 1 may also include a decompression unit 15 that decompresses the compressed data and combines uncompressed upper bit data or lower bit data stored in the storage unit 14 in association with the compressed data. .

Furthermore, the data compression apparatus 1 includes an arithmetic control unit 16 such as a CPU (Central Processing Unit) that controls the operation of each unit, and mainly functions of a dividing unit 11, a determining unit 12, a compressing unit 13, and a restoring unit 15. Execute.
Here, the arithmetic control unit 16 activates an OS or the like stored in the storage unit 14, reads and executes a program stored on the OS, and controls each unit or the entire apparatus, thereby providing a data compression apparatus. Perform unique actions.
At this time, the storage unit 14 expands the program or application executed by the arithmetic control unit 16 in the program storage area, and temporarily stores the input data or data such as processing results generated when the program or application is executed in the work area. You may memorize in.

The data compression apparatus 1 of the present invention may include a communication unit 17 that is electrically connected to an external device via a connection line, a network, or the like.
The data compression apparatus 1 may transmit the compressed data compressed by the compression unit 13 to an external device via the communication unit 17.

(Description of main components)
When the dividing unit 11 determines the “divided bit number” as 8 bits in advance, the pixel value (16-bit data) for one pixel of the image data is an example of upper 8 bit data and lower bit data as an example of upper bit data. Are divided into lower 8-bit data.
In this case, the dividing unit 11 divides the pixel value of one pixel composed of N (N is a natural number) bit data into an N-bit upper N / 2 bit portion and an N-bit lower N / 2 bit portion. Become.

As shown in FIG. 2, the dividing unit 11 divides the 16 bits of each pixel into upper 8 bits and lower 8 bits and stores them in the storage unit 14 as two 8-bit one-dimensional arrays.
Incidentally, the lower 8 bits are 0 ² pixel value as a component, ... represent ^{2 7,} upper 8 bits represent the ² 8, ^{..., 2 15.}

Here, the dividing unit 11 may be configured to simply divide the boundary between the upper bits and the lower bits into half (N / 2) by setting the “number of divided bits” to 8 bits in advance, but is not limited thereto.
For example, the dividing unit 11 calculates a statistical value (average and standard deviation, etc.) of the background distribution from the histogram of pixel values, for example, x times the standard deviation than the average (x sigma (for example, x is a numerical value of about 2 to 3). )) A bit area corresponding to a lower pixel value may be a lower bit. In this way, it is possible to efficiently increase the compression rate on the upper bit side.

The determining unit 12 determines whether at least one of the upper 8 bit data and the lower 8 bit data divided by the dividing unit 11 is suitable for compression.
Specifically, the determination unit 12 performs at least one of the following compression processes on pixel values of a plurality of predetermined pixels in the image data, and compresses the compression rate (the amount of data after compression). ) Divided by the amount of data before compression), based on at least one of the compression times, the compression processing with the calculated compression rate is the minimum compression rate or the compression processing with the shortest measured compression time It is determined that at least one of the upper 8-bit data and the lower 8-bit data extracted and compressed by the extracted compression process is suitable for compression.
(1) Compression of upper 8 bits data only (hereinafter referred to as compression processing (1))
(2) Compression of only lower 8 bits data (hereinafter referred to as compression process (2))
(3) Compression of the entire bit data (16-bit data) for one pixel (hereinafter referred to as compression processing (3))

  Here, since the determination unit 12 may compress the lower bits depending on the characteristics of the compressed image data, the determination unit 12 performs the compression process (2) including the lower bits. The determination unit 12 does not necessarily have to perform the compression process (2).

Further, the determination means 12 may further perform a compression process (hereinafter referred to as compression process (4)) of “compress both the higher-order 8-bit data and the lower-order 8-bit data after division”.
At this time, if the data after the compression process (4) becomes smaller than before compression, the determination unit 12 determines that the upper 8 bit data and the lower 8 bit data are suitable for compression, and does not become smaller than before compression. In some cases, it may be determined that it is not suitable for compression. This is because some compression algorithms become larger than the original data when compression is attempted. Further, in consideration of the time required for transferring and decompressing the compressed data, it may be determined whether the cost of the compression operation is commensurate with the performance improvement, and whether to perform the compression is determined.

When at least one of the divided upper 8 bit data and lower 8 bit data is suitable for compression, the compression unit 13 selects all the pixels (or some of the pixels) of the image data captured in the same wavelength band as the image data. ) At least one of the upper 8-bit data and the lower 8-bit data.
For example, when the upper 8 bit data is suitable for compression, the compression unit 13 collects the upper 8 bit data in all pixels (or a part of the pixels) of the image data captured in the same wavelength band as the image data. Compress.
The compression unit 13 stores the compressed bit data in the storage unit 14 and records which data is compressed.

The compression algorithm used by the compression means 13 may be a general-purpose lossless compression algorithm (Zip, Deflate, Gzip, Bzip2, LZW, simple Run Length compression, or the like).
Further, when compressing at least one of the upper 8 bit data and the lower 8 bit data using a plurality of compression algorithms, the compression unit 13 records which algorithm is used for compression.

The storage unit 14 stores image data captured in a plurality of types of wavelength bands acquired from an external device, and stores the compressed data compressed by the compression unit 13. Further, the storage unit 14 may store “image type information” (for example, information for each type of wavelength band) in which image data and a type of image data (for example, a wavelength band taken in the image data) are associated. .
When the compression unit 13 compresses only the upper 8 bit data or the lower 8 bit data, the storage unit 14 stores the compressed data and the uncompressed lower 8 bit data or the upper 8 bit data in association with each other.

  The storage means 14 is, for example, a RAM (Random Access Memory), a ROM (Read Only Memory), a hard disk device, or the like, and a program or application for operating mainly as the OS or the data compression device 1 or when executing these programs. Various information such as data to be used is stored.

The restoring unit 15 restores the compressed data.
Specifically, when the compression unit 13 compresses only the upper 8 bit data or the lower 8 bit data, the decompression unit 15 stores the uncompressed lower 8 bit data stored in the storage unit 14 in association with the compressed data. Alternatively, the upper 8 bit data and the restored upper 8 bit data or lower 8 bit data are combined to restore the compressed data.

(Main features of the present invention)
The main feature of the present invention is that the dividing means 11 divides the data into upper bit data and lower bit data using the characteristics and properties of the image data.

Here, the image data includes an image taken for measurement.
For example, the image data includes a “microscope cell image” and a “telescope image”.
Specifically, a sample of a cell (administered with a drug) is irradiated with excitation light, and a cell for drug discovery screening by image processing based on a fluorescence signal from the cell is photographed in a plurality of types of wavelength bands. The obtained image data is given as an example (image example in FIG. 5 described later).

In such image data, in a region where an object to be measured is included in the field of view of the image, the gray value is light and the pixel value is high, and it can be expressed that both the lower 8 bits and the upper 8 bits are not used. Absent.
On the other hand, in the background area of the image field of view where the gray scale density is dark, it can be expressed by the lower 8 bits.

In addition, the lower 8-bit data expresses the random nature of background noise (such as readout noise or dark current caused by the image sensor) and signal noise, and is therefore often less regular and difficult to compress.
On the other hand, the high-order 8-bit data is a bright signal area (the gray scale is light and shaded), and the data regularity is high especially when the ratio of the data in the screen is relatively small. Can be realized.

  In general, in an imaging device (such as a CCD (Charge Coupled Device) device) that performs high-accuracy measurement, when the distribution of noise pixel values in a state in which light is not irradiated is AD (analog-digital) converted, the noise width (for example, The gain is adjusted so that sampling is performed with about 10 channels or more as the standard deviation) as an AD conversion value. This is to reduce the quantization error due to AD conversion.

For this reason, in order to express noise, information of at least about 5 to 6 bits or more is required, so about 5 to 6 bits (or more) of the lower 8 bits data is highly random, and reversible compression is extremely difficult. . (In general, with any compression algorithm, highly regular data is easy to compress, and random data is difficult to compress.)
For these reasons, if this noise component is compressed by a general-purpose reversible compression algorithm, a high compression rate cannot be obtained for a long CPU load time.

  As described above, the main feature of the present invention is that the dividing unit 11 divides the bit data into the upper bit data and the lower bit data using the features and characteristics of the image data as described above, and the bit data divided by the determining unit. When at least one of the bit data divided by the compression means is suitable for compression, image data of the same type or similar to the image data (for example, substantially the same or By compressing at least one of the upper bit data and lower bit data of image data shot in the same wavelength band, etc., there is less waste in computation than conventional compression techniques (to achieve an equivalent compression ratio) The amount of calculation is small and high speed), and efficient compression can be performed.

(Explanation of layout and connection of main components)
The dividing unit 11 is connected to the storage unit 14 and reads image data, and divides a pixel value (N-bit data) for one pixel into upper bit data and lower bit data. The divided bit data is stored in the storage unit 14.
The determination unit 12 is connected to the storage unit 14, reads pixel values of a plurality of predetermined pixels from the image data, performs compression by a predetermined method, and determines whether or not the compression is suitable.

The compression unit 13 is connected to the storage unit 14 and, when it is determined that the compression unit 13 is suitable for compression, is the same type or similar image data as the image data (for example, image data photographed in substantially the same or the same wavelength band). At least one of upper bit data and lower bit data of all pixels (or a part) of the captured image data is read and compressed. The compressed data is stored in the storage unit 14.
At this time, when the compression means 13 compresses only the upper bit data or the lower bit data of all the pixels (or a part), the compression means 13 associates the compressed data with the lower bit data or the upper bit data which is uncompressed and stores the storage means 14. To remember.

  The restoring unit 15 is connected to the storage unit 14 and reads the compressed data to restore the data. At this time, when only the upper or lower bit data is compressed, the compressed data is restored by combining with the uncompressed bit data stored in the storage unit 14 in association with the compressed data. The restored data may be stored in the storage unit 14.

(Description of operation)

With such a configuration, the data compression apparatus of the present invention performs the following operation. FIG. 3 is an operation flow diagram in which the data compression apparatus of FIG. 1 compresses data related to a large number of images.
In step SP101 in FIG. 3, the dividing unit 11 reads (first sheet) image data from the storage unit 14.

In step SP102, the dividing unit 11 divides the pixel value for one pixel into upper 8-bit data and lower 8-bit data, and stores the two 8-bit data in the storage unit 14 as a one-dimensional array.
This operation is performed on pixel values (bit data) of a plurality of predetermined pixels or all pixels.

  In step SP103, the determination unit 12 performs compression evaluation to determine whether at least one of the upper 8-bit data and the lower 8-bit data divided by the dividing unit 11 is suitable for compression, and determines the type of the image data ( For example, “information to be compressed by image type” (for example, information to be compressed by wavelength) in which information indicating the type of wavelength band at the time of photographing) and information indicating bit data to be compressed having compression compatibility is stored. Store in the means 14.

Specifically, the determination unit 12 performs (1) compression processing of only the upper 8-bit data (hereinafter referred to as compression processing (1)), (2) for pixel values of a plurality of predetermined pixels in the image data. ) Each of compression processing (hereinafter referred to as compression processing (2)) of only lower 8-bit data, and (3) compression processing (hereinafter referred to as compression processing (3)) of the entire bit data (16-bit data) for one pixel. A compression process is performed, and at least one of each compression rate (a ratio obtained by dividing the amount of data after compression by the amount of data before compression) and the compression time are calculated and measured.
The determination unit 12 extracts the compression process in which the calculated compression ratio is the minimum compression ratio or the compression process in which the measured compression time is the shortest time, and the upper 8-bit data compressed by the extracted compression process or It is determined that at least one of the lower 8-bit data is suitable for compression. The determination unit 12 stores, in the storage unit 14, “compression target information for each image type” in which information indicating the imaging wavelength band of the image data is associated with information indicating compression-compatible bit data to be compressed. To do.

FIG. 5 is an explanatory diagram of the determination operation by the determination unit 12 of FIG. 1, (A) is an example of image data used for compression evaluation, and (B) is a description of the compression rate in each compression process (1) to (3). Table (C) is a graph of (B).
Here, the compression algorithm used by the compression unit 11 is exemplified by a compression result using a general-purpose GZipStream.

  In (A), the images used for the compression evaluation (numbered 1, 2, and 3 in order from the left) are shown. The sample of the cell to which the drug was administered was irradiated with excitation light, and based on the fluorescence signal from the cell. It is the image data which image | photographed the cell for performing a drug discovery screening by image processing in several types of wavelength bands. The three images are taken in the same field of view and in different wavelength bands.

In (B), the compression rate is a ratio obtained by dividing "data amount after compression" by "data amount before compression". The smaller the better, the better. Means that.
The image number in the table indicates the image number of (A). “1 column as a whole” indicates a compression rate when the entire image is compressed as a 1-byte byte array without dividing bit data for one pixel.

  “Lower bits” and “Upper bits” indicate the compression rates when the two columns of arrays divided by this method are individually compressed. In these two columns, the denominator of the compression rate is the respective data amount (half of the whole). In the “lower bits” and “upper bits”, only the upper bits are clearly superior in compression rate.

"Both columns and both compression" indicates the compression ratio of the entire data when both the upper 8 bits and the lower 8 bits after division are compressed, and corresponds to the compression process (3).
“Compression of upper bits only” indicates the compression ratio of the entire data when the upper bits are compressed and the lower bits are not compressed, and corresponds to the compression process (1).

(C) is a graph of the data in the table of (B), where the horizontal axis corresponds to the image number, and each of the five bar graphs corresponds to the column of (B).
However, in order to make it easy to understand each contribution when the total data amount is 1.0, the compression rate of “lower bits” and “upper bits” is different from the values in the table, and the denominator is set for each data amount. Instead, it is displayed as a ratio to the total amount of data, which is half the value in the table of (B).

In the case of the compression ratios as in (B) and (C) described above, the determination unit 12 uses the “compression of upper bits only” compression ratio (0.560 for image number 1) as “the whole is 1. Column "(0.976) and" compress both in two columns "(0.867). Therefore, the lower bits are judged not to be compressed, and only the upper bits are suitable for compression. judge.
If compression is performed by the method of compressing only the upper bits, the compression rate is excellent and the lower bits are not compressed. Therefore, the CPU resource (calculation time) required for compression can be reduced to about half, which is effective in terms of high efficiency.

  As described above, the determination unit 12 performs compression evaluation to determine whether at least one of the upper 8 bit data and the lower 8 bit data divided by the dividing unit 11 is suitable for compression, and the image data is captured. The storage unit 14 stores “compression target information for each image type” in which information indicating a wavelength band and information indicating compression-compatible bit data to be compressed are associated with each other.

Note that when the compression ratio of “compress both in two columns” is the minimum as a result of the compression evaluation, the determination unit 12 determines that both the upper 8 bits and the lower 8 bits are suitable for compression.
For the sake of simplicity, in this table, the compression ratio of the entire data when only the lower 8 bits are compressed and the upper 8 bits are not compressed (in the case of compression processing (2)) is omitted. In this case, when the compression rate of the compression process (2) is the minimum as a result of the compression evaluation, the determination unit 12 determines that the lower 8 bits are suitable for compression.

  In step SP104, when the compression unit 13 determines that at least one of the upper 8 bit data and the lower 8 bit data divided by the determination unit 12 is suitable for compression, the image is captured in the same wavelength band as the image data. Compress at least one of upper 8 bit data and lower 8 bit data of the data.

  For example, when the determination unit 12 determines that only the upper 8-bit data is suitable for compression because the determination unit 12 indicates that the compression process (1) has the best compression rate as described above, Only the upper 8-bit data of all the pixels (partial pixels) is compressed and stored in the storage means 14.

  In step SP105, the arithmetic control unit 16 reads the compressed data from the storage unit 14, and controls the communication unit 17 to transmit the compressed data for one image to an external device.

  In step SP106, the arithmetic control unit 16 determines whether or not other image data before compression is stored in the storage unit 14, and if other image data is stored, the operation control unit 16 proceeds to step SP107. If no image data is stored, the process ends.

  In step SP107, the dividing unit 11 reads the image data (image data before compression) from the storage unit 14, and divides the pixel value for one pixel into upper 8 bit data and lower 8 bit data. Bit data is stored in the storage means 14 as a one-dimensional array. This operation is performed on pixel values (bit data) of a plurality of predetermined pixels or all pixels.

In step SP108, the compression means 13 is based on the “image type information” and “compression target information for each image type” of the image data stored in the storage means 14, and is the same type as the image data (for example, photographing the image data). The presence or absence of bit data having compression suitability for the selected wavelength band).

When there is no bit data having compression compatibility of the same type as the image data (for example, a wavelength band taken by the image data) in the “compression target information by image type” in the data compression apparatus of the present invention, the step The process proceeds to SP102.
The data compression apparatus of the present invention includes a step when the “compression target information for each image type” includes bit data having compression compatibility of the same type as the image data (for example, a wavelength band taken by the image data). The process proceeds to SP109.

In step SP109, the compression means 13 indicates the bit data having compression suitability in the same type (for example, a wavelength band taken by the image data) as the image data indicated in the “compression target information by image type”. Compressed and stored in the storage means 14.
For example, the compression means 13 shows the compression method (1) that the compression means (1) is the best (minimum) compression rate by the determination means 12 in SP103 for the compression method of image data shot in the same wavelength band as the image data. When it is determined that only the upper 8 bit data is suitable for compression, only the upper 8 bit data is compressed and stored in the storage means 14.

  That is, when it is determined that at least one of the upper 8-bit data and the lower 8-bit data divided by the determination unit 12 is suitable for compression, the compression unit 13 is the same type of image data as the image data (for example, the image The upper 8 bit data or the lower 8 bit data of the image data captured in the same wavelength band as the data is compressed.

  In step SP110, the arithmetic control unit 16 reads the compressed data from the storage unit 14, and controls the communication unit 17 to transmit the compressed data for one image to an external device.

  In step SP111, the arithmetic control unit 16 determines whether or not other image data before compression is stored in the storage unit 14, and if other image data is stored, the operation control unit 16 proceeds to step SP107. If no image data is stored, the process ends.

The data compression apparatus of the present invention also performs the following operation. FIG. 5 is an operation flow chart when the data compression apparatus of FIG. 1 compresses data (for only one image).
In step SP151 in FIG. 5, the dividing unit 11 reads the image data (image data before compression) from the storage unit 14, and divides the pixel value for one pixel into upper 8 bit data and lower 8 bit data, Two 8-bit data are stored in the storage means 14 as a one-dimensional array. This operation is performed on pixel values (bit data) of a plurality of predetermined pixels or all pixels.

In step SP152, based on the “image type information” and “compression target information by image type” of the image data stored in the storage unit 14, the compression unit 13 is the same type as the image data (for example, photographing the image data). The presence or absence of bit data having compression suitability for the selected wavelength band).

If there is no bit data having compression compatibility of the same type as the image data (for example, a wavelength band taken by the image data) in the “image type compression target information”, the process proceeds to step SP153.
If there is bit data having compression compatibility of the same type as the image data (for example, a wavelength band taken by the image data) in the “image type compression target information”, the process proceeds to step SP154.

  In step SP153, the determination unit 12 performs compression evaluation to determine whether at least one of the upper 8 bit data and the lower 8 bit data divided by the dividing unit 11 is suitable for compression, and determines the type of the image data ( For example, storage means 14 stores “compression target information by image type” (for example, compression target information by wavelength) in which information indicating, for example, the imaging wavelength band) and information indicating compression-compatible bit data to be compressed are associated. (The same operation as SP103 described above) is performed.

In step SP154, the compression means 13 indicates the bit data having compression suitability in the same type (for example, a wavelength band taken by the image data) as the image data indicated in the “compression target information by image type”. Compressed and stored in the storage means 14.
For example, the compression means 13 shows the compression method (1) that the compression means (1) is the best (minimum) compression rate by the determination means 12 in SP103 for the compression method of image data shot in the same wavelength band as the image data. When it is determined that only the upper 8 bit data is suitable for compression, only the upper 8 bit data is compressed and stored in the storage means 14.

  That is, when it is determined that at least one of the upper 8-bit data and the lower 8-bit data divided by the determination unit 12 is suitable for compression, the compression unit 13 is the same type of image data as the image data (for example, the image The upper 8 bit data or the lower 8 bit data of the image data captured in the same wavelength band as the data is compressed.

  In step SP155, the arithmetic control unit 16 reads the compressed data from the storage unit 14, controls the communication unit 17, and transmits the compressed data to an external device.

  As a result, the data compression apparatus of the present invention uses the characteristics and properties of the image data to divide the dividing means 11 into upper bit data and lower bit data, and compress at least one of the bit data divided by the judging means. If at least one of the bit data divided by the compression means is suitable for compression, the image data is the same or similar to the image data (for example, photographed in substantially the same or the same wavelength band). By compressing at least one of the upper bit data and lower bit data of the image data), there is no waste in calculation compared to the conventional compression technology (the calculation amount for achieving the same compression ratio is small and high speed). Yes), efficient compression can be performed.

  In addition, since the data compression apparatus of the present invention operates in units of 8 bits (1 byte), it is effective in that it can be easily increased in speed by utilizing a bit shift operation in programming.

<Second embodiment>
In addition to the configuration described in the first embodiment, the data compression apparatus according to the present invention provides each pixel as preprocessing for dividing the pixel value of one pixel of image data into upper 8 bit data and lower 8 bit data. Subtracting means 17 for subtracting the minimum pixel value in the image data from the value may be provided. In this case, the arithmetic control unit 16 also executes the function of the subtracting means 17.

  As a result, the data compression apparatus according to the present invention is also provided with the subtracting means 17 so that the offset of the entire image can be shifted to the low bit side by subtracting the minimum value of the data. Can be efficiently shifted (pressed) to the lower bit side, and the compression rate on the upper bit side can be improved.

  The data compression device 1 or the data restoration device 2 separately stores the minimum pixel value or offset value, and adds the minimum pixel value or offset value to the previously stored image data restored by the restoration unit 15 or the restoration unit 22. A reversible compression may be realized by addition.

<Third embodiment>
In the first embodiment, the dividing unit 11 of the data compressing apparatus 1 simply divides the image data for one pixel into upper 8 bit data and lower 8 bit data to obtain two 8-bit one-dimensional data. Although described as being stored in the storage means 14 as an array, the upper 8 bit data and the lower 8 bit data are divided into a plurality of data arrays corresponding to “rows” and “columns” of the two-dimensional image data. 14 may be stored.

  Specifically, the dividing means 11 of the data compression apparatus of the present embodiment divides the image data into a plurality of data arrays corresponding to the rows or columns, and each pixel in the row or column for each row or column. The value may be divided and stored into upper 8 bit data and lower 8 bit data.

In addition, the determination unit 12 of the data compression apparatus according to the present embodiment selects a plurality of representative rows or columns, and compresses at least one of the upper bit data and the lower bit data for each selected row or column ( Perform at least one of 1) to (4).
At this time, the determination unit 12 extracts a compression process having a compression rate larger than a predetermined threshold from these compression processes, and compresses the upper bit data, the lower bit data, and the undivided bits compressed by the extracted compression process. It is determined that at least one of the bit data is incompatible with compression.

  Furthermore, when it is determined that at least one of the upper 8 bit data and the lower 8 bit data is suitable for compression, the compression unit 13 of the data compression apparatus of the present embodiment is taken in the same wavelength band as the image data. For image data, at least one of upper 8 bit data and lower 8 bit data determined to be suitable for compression is compressed for each “row” or “column”.

  As a result, in the data compression apparatus of the present invention, the dividing unit 11 divides the image data into a plurality of data arrays corresponding to the rows or columns, and each pixel value in these rows or columns is converted into upper 8-bit data and lower-order data. By dividing the data into 8-bit data, the compression unit 13 can compress each one-dimensional array in parallel, which is effective in that the speed can be easily increased.

  In addition, the determination unit 12 of the data compression apparatus of the embodiment selects a plurality of representative rows for the upper 8 bit data and the lower 8 bit data divided for each row, and the upper (lower) 8 of the selected row. By trying to compress bit data, the overall compression rate is predicted, and when it is determined that a compression rate smaller than a predetermined value cannot be obtained (an excellent compression rate cannot be obtained), or in advance When a compression rate larger than a predetermined value is obtained, it is determined that the upper (lower) 8-bit data is not suitable for compression.

  For this reason, the data compression apparatus according to the present embodiment compresses bit data for each of a plurality of rows (columns) selected by the determination unit 12 and determines all pixels (partially) as described in the first embodiment. Compared with the case where compression is performed on the upper 8-bit data or lower-order 8-bit data of the pixel) to determine whether each bit data is suitable for compression, it is possible to reduce wasteful calculation and to make a quick determination. It is valid.

<Fourth embodiment>
In the data compression apparatuses of the first to third embodiments, a two-dimensional image is handled as an array of one column (recording the number of rows and columns separately). That is, the scanning order of the pixels is, for example, after scanning one row (or one column) with position coordinates such as (0, 0), (0, 1), (0, 2),. However, there is a possibility that a higher compression ratio can be realized by changing the scanning method.
For example, in an image showing an object such as a cell, since the signal of the object is a block, the block is slightly smaller than the block (for example, 16 × 16, 32 × 32,. .. When scanning is performed in units of pixel squares or the like, regions where signals exist are likely to appear continuously. That is, there is a possibility that the regularity on the upper bit side is increased and the compression rate can be improved.

  In view of these points, the data compression apparatus of the present invention is not limited to the function of the dividing means 11 of <First Embodiment> to <Third Embodiment>. A block composed of a plurality of blocks, each of which is smaller than the range occupied by one unit to be measured, and each pixel value in these blocks is divided into upper 8-bit data and lower 8-bit data. Also good.

  Specifically, the dividing unit 11 of the data compression apparatus of the present embodiment divides the image data into a plurality of blocks each of which is a block made up of a plurality of pixels and smaller than the range occupied by one unit of the measurement target, Each pixel value in these blocks may be divided into upper 8 bit data and lower 8 bit data.

  In addition, the determination unit 12 of the data compression apparatus according to the present embodiment performs at least one of the compression processes (1) to (4) for at least one of the upper 8 bit data and the lower 8 bit data for each of a plurality of predetermined blocks. Do something. The determination unit 12 determines whether at least one of the upper 8-bit data and the lower 8-bit data is suitable for compression.

  Further, the compression means 13 of the data compression apparatus of the present embodiment is photographed in the same wavelength band as the image data when it is determined that at least one of the upper 8 bit data and the lower 8 bit data is suitable for compression. For image data, at least one of upper 8 bit data and lower 8 bit data determined to be suitable for compression for each block is compressed.

  As a result, the dividing unit 11 divides the image data into a plurality of “blocks” each of which is a block made up of a plurality of pixels and is smaller than the range occupied by one unit of the measurement target. The image data is divided into upper 8 bit data and lower 8 bit data, and the determination unit 12 determines whether or not the image data is suitable for compression for each “block”, and the compression unit 13 captures images in the same wavelength band as the image data. By compressing the processed image data for each “block”, the regularity on the upper bit side is increased, and the compression rate can be improved.

In addition, the data compression apparatus according to the present embodiment is considered to be effective particularly when the type and outline of the imaging target are known in advance.
Further, the data compression apparatus of the present embodiment may be variable so that the block size becomes an optimum value according to the photographing magnification or the like as long as it is a cell image.

<Fifth embodiment>
In the above-described embodiment, the data compression apparatus 1 according to the present invention is described as having the restoration unit 15. However, the present invention is not limited to this, and a data restoration apparatus having a restoration function is provided. The data compressed by the data compression device 1 may be restored.

FIG. 6 is an explanatory diagram of the configuration of an embodiment of the data restoration apparatus of the present invention.
In FIG. 6, the data restoration device 2 mainly includes a storage unit 21 and a restoration unit 22, an arithmetic control unit 23 such as a CPU for controlling the operation of each unit, an external device (for example, the data compression device 1) and a connection line. And communication means 24 electrically connected via a network or the like.

  The storage unit 21 stores the compressed data received from the data compression device 1 via a network or the like. Specifically, when only one of the upper 8 bit data and the lower 8 bit data divided by the data compression apparatus 1 is compressed, the storage unit 21 compresses the compressed data and the lower bit data that is not compressed or the upper bit data. Store bit data in association with each other.

The restoring unit 22 restores the compressed data stored in the storage unit 21.
In addition, the decompressing unit 22 is configured to store the uncompressed lower level stored in the storage unit 21 in association with the compressed data when the compression unit 13 of the data compression apparatus 1 compresses only the upper 8 bit data or the lower 8 bit data. The compressed data is restored by combining the 8-bit data or the upper 8-bit data with the restored upper 8-bit data or the lower 8-bit data.

FIG. 7 is an operation flowchart when the data decompression apparatus of the present invention decompresses (decompresses) compressed data.
In step SP201 in FIG. 7, the restoring unit 22 reads the compressed data from the storage unit 21, and restores the compressed upper 8 bit data or lower 8 bit data among the divided data.

  In step SP202, the decompressing unit 22 restores the restored upper or lower 8 bit data and the uncompressed lower 8 bit data or the lower 8 bits stored in the storage unit 21 in association with the upper or lower 8 bit data. Combine the data and restore (generate) the original data string.

  In step SP203, the arithmetic control unit 23 controls the communication unit 24 to transmit the restored data to other devices via the network.

  As a result, when the data compression apparatus compresses only one of the upper bit data and the lower bit data divided by the data compression apparatus at a predetermined number of divided bits smaller than N bits, the compressed data And storage means for storing the uncompressed lower bit data or upper bit data in association with each other, and decompressing the compressed data and storing the compressed data in the storage means in association with the compressed data By providing the restoring means for combining the two, calculation is less wasteful than the conventional technique (the calculation amount for achieving the same compression rate is small and high speed), and efficient restoration can be performed.

<Sixth embodiment>
The data compression apparatus 1 of the present invention may be used in an image analysis system that performs image analysis. FIG. 8 is a structural explanatory view showing an embodiment of an image analysis system using the data compression apparatus 1 of FIG. 1. The same reference numerals are given to the parts common to FIG. 1, and the description will be omitted as appropriate.

  In FIG. 8, an image analysis system 50 irradiates a sample of a cell to which a drug is administered with excitation light, performs image processing based on a fluorescence signal from the cell, and performs drug discovery screening. An image that is captured in a wavelength band and that is configured by a camera system 51 that transmits image data obtained by capturing, and a general-purpose computer or the like, acquires image data from the camera system 51, compresses the acquired data, and transmits compressed data The acquisition system 52, a network device 53 such as a switch for transferring the compressed data from the image acquisition system 52, and the compressed data acquired via the network device 53 are stored (the compression means 13 stores the upper 8 bit data or the lower 8 bits. When only data is compressed, this compressed data and uncompressed lower 8-bit data or upper 8 In association with the Ttodeta) and the storage system 54, and the compressed data to obtain restored reduction from an image analysis device 55 performs image analysis from the storage system 54.

  Alternatively, the camera system may execute the data compression method provided by the data compression apparatus 1 of the present invention with the above-described configuration. In this case, the camera system may include the above-described dividing unit 11, determination unit 12, compression unit 13, and storage unit 14, and may transmit compressed data to the image acquisition system 52.

<Other examples>
The data compression apparatus according to the present invention has been described as an operation on an image once stored as 16 bits in each of the above-described embodiments. However, an area close to a camera (a DSP, FPGA, CPU in the camera), a camera capture board, or the like. A compression function may be mounted on the above hardware (DSP, FPGA, CPU) and output in a compressed format when output from the camera.

The data compression apparatus of the present invention implicitly assumes that the divided upper and lower bit data are held as one data (class, structure or file) in each of the embodiments described above, but the present invention is not limited to this. Rather than a single image including two pieces of 8-bit image data, it may be as if two pixels of 8-bit images were acquired.
This is effective in that the present invention can be applied without using a special data format. In other words, a method of compressing only an image corresponding to the upper bits in an image of 8 bits per pixel by a general-purpose method can be considered.

As described above, the data compression apparatus of the present invention pays attention to the characteristics of the image data (mainly the lower bit area representing the noise component and the upper bit area mainly representing the signal component), and uses this to divide means 11. Is divided into upper bit data and lower bit data, and determination means determines whether at least one of the divided bit data is suitable for compression, and at least one of the bit data divided by compression means is suitable for compression When compressing the image data, at least one of the upper bit data and the lower bit data of the image data captured in the same wavelength band as the image data is compressed. This is effective in that the amount of calculation for achieving the rate is small and the speed is high), and efficient compression can be performed. Thereby, useless calculation time can be reduced, and efficient image data compression can be realized.
(Appendix 8)
The image data dividing means includes
7. The data compression apparatus according to claim 1, wherein the pixel value for one pixel is divided into an upper N / 2 bit portion of N bits and a lower N / 2 bit portion of N bits. .

DESCRIPTION OF SYMBOLS 1 Data compression apparatus 11 Division | segmentation means 12 Determination means 13 Compression means 14 Storage means 15 Restoration means 16 Operation control part

Claims (7)

Image data obtained by photographing the measurement object in one or more types of wavelength bands, and reversibly converted to image data composed of a series of pixel values for one pixel, which is a numerical value represented by a predetermined N (N is a natural number) bit number. In a data compression apparatus that performs compression processing,
Image data dividing means for dividing the pixel value for one pixel into upper bit data and lower bit data at a predetermined number of divided bits smaller than the N bits;
Determining means for determining whether at least one of the divided bit data is suitable for compression;
A compression unit that compresses at least one of the upper bit data or the lower bit data of the same or similar image data to the image data when at least one of the divided bit data is suitable for compression; A data compression apparatus. The determination means includes
A compression process in which at least one of the following (1) to (5) is performed on pixel values of a plurality of predetermined pixels in the image data, and the calculated compression ratio is the minimum compression ratio or A compression process in which the measured compression time is the shortest time is extracted, and it is determined that at least one of the upper bit data and the lower bit data compressed by the extracted compression process is suitable for compression. The data compression apparatus according to claim 1.
(1) Compression of only the upper bit data (2) Compression of only the lower bit data (3) Compression of undivided bit data (4) Compression of the upper bit data and the lower bit data (5) Division or not No compression of divided bit data When the compression means compresses only the upper bit data or the lower bit data, the compression means comprises storage means for storing the compressed data and the lower bit data or the upper bit data that are uncompressed in association with each other,
The image data dividing means is
Before the pixel value for one pixel is divided into the upper bit data and the lower bit data, subtracting means for subtracting the minimum value of the pixel value in the image data from each pixel value, The data compression apparatus according to claim 1 or 2. The image data dividing means divides the image data into data arrays corresponding to rows or columns, divides each pixel value in these rows or columns into the upper bit data and the lower bit data,
The determination means performs at least one of the compression processes (1) to (5) for at least one of the upper bit data or the lower bit data for each of a plurality of predetermined rows or columns. A compression process having a compression ratio larger than a predetermined threshold is extracted from the compression processes of the above, and at least one of the upper bit data, the lower bit data, and the undivided bit data compressed by the extracted compression process Is determined to be incompatible with compression,
When it is determined that at least one of the upper bit data or the lower bit data is suitable for compression, the compression means is adapted to compression for each row or column of the same or similar image data as the image data. The data compression apparatus according to claim 2 or 3, wherein at least one of the determined upper bit data or the lower bit data is compressed. The image data dividing means divides the image data into blocks each composed of a plurality of pixels and smaller than a range occupied by one unit of a measurement target, and sets each pixel value in these blocks Divided into bit data and the lower bit data,
The determination means performs at least one of the compression processes (1) to (5) for at least one of the upper bit data and the lower bit data for each predetermined block,
When it is determined that at least one of the upper bit data and the lower bit data is suitable for compression, the compression unit determines that the same or similar image data as the image data is suitable for compression for each block. 4. The data compression apparatus according to claim 2, wherein at least one of the upper bit data and the lower bit data is compressed. Reversible compression processing is performed on image data captured in one or a plurality of wavelength bands and composed of a series of pixel values for one pixel, which is a numerical value represented by a predetermined N (N is a natural number) bit number. In the data decompression device which decompresses the compressed data compressed by the data compression device according to any one of claims 1 to 3,
Stores compressed data of image data compressed by the data compression device, and only one of upper bit data and lower bit data divided by the data compression device at a predetermined division bit number smaller than the N bits Storage means for storing the compressed data and the uncompressed lower bit data or the upper bit data in association with each other,
A data decompression device comprising decompression means for decompressing the compressed data and combining the uncompressed upper bit data or the lower bit data stored in the storage means in association with the compressed data .   The image data is image data obtained by irradiating a cell sample with excitation light, imaging the cell for drug discovery screening by performing image processing based on a fluorescence signal from the cell, in a plurality of types of wavelength bands. The data compression device according to any one of claims 1 to 5, or the data restoration device according to claim 6.