JP2003242501A - Image processing apparatus and method, image forming apparatus, and computer-readable program - Google Patents

Abstract

(57) [Problem] To provide an image processing technique capable of effectively removing noise from dark current noise while suppressing the processing amount at the time of noise removal. SOLUTION: Storage means for storing a correspondence relationship between luminance information and a size of a noise removal filter, and for each pixel of image data, selecting a size corresponding to the luminance information of the pixel by referring to the correspondence relationship. Means for performing a noise removal process using the noise removal filter of the selected size. The correspondence is determined so that luminance information indicating lower luminance corresponds to a larger size.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an image processing technique for removing noise from image data, and more particularly to a noise removing technique suitable for execution in an image forming apparatus such as a printer.

[0002]

2. Description of the Related Art In recent years, the rapid spread of digital cameras and the like,
Along with the price reduction of printers, the number of users who enjoy printing images taken with a digital camera or the like on a printer at home is increasing. In such a case, printing was generally performed via a personal computer in the past, but recently, images are directly read from a digital camera (or its flash memory for recording) without using a personal computer. Printers that can print by using a printer (hereinafter, referred to as "direct printer") have begun to appear.

[0003]

When an image is taken with a digital camera or the like, image deterioration due to the characteristics of the image sensor poses a problem.

For example, an image sensor such as CCD or CMOS used in a digital camera or the like has a characteristic that electric charges are generated by heat even when the light receiving element is not exposed to light, and such electric charges are dark current noise ( This causes deterioration of image quality called thermal noise). In particular, CMOS image sensors are often used in low-priced digital cameras, and CMOS image sensors are C
Since the sensitivity is generally lower than that of a CD, it tends to be easily affected by image quality deterioration due to dark current noise.

When such image deterioration has occurred, the image is printed with high quality (an image is formed).
In order to do so, it is important to apply noise removal processing using an appropriate filter.

Here, in the case of printing through the host device, the above-mentioned image processing can be performed at high speed in the host device having a high processing capacity, but in the case of the direct printer, the printer itself executes these processes. It is necessary to perform the image processing of.

However, a CP that is generally installed in a printer
U has a lower processing capacity than a personal computer, has less cache, and is slow in memory access. Therefore, U is simply configured to perform filtering processing that needs to refer to pixel values to perform noise removal. If so, there arises a problem that the printing speed is significantly reduced.

In view of this, the present invention provides an image processing technique capable of effectively removing noise while suppressing the processing amount when removing noise, and using such image processing technique, the printing speed is improved. An object of the present invention is to provide an image forming apparatus such as a printer that can perform high-quality printing without dropping it unnecessarily.

[0009]

An image processing apparatus according to the present invention comprises a storage unit for storing a correspondence relationship between luminance information and a size of a noise removal filter, and a pixel for each pixel of image data.
A unit for selecting a size corresponding to the luminance information of the pixel with reference to the correspondence relationship and executing a noise removal process using the noise removal filter of the selected size.

It is desirable that the correspondence relation is determined so that the luminance information indicating lower luminance corresponds to a larger size.

The image forming apparatus of the present invention is the image forming apparatus according to claim 1 or 2.
The image processing device described above is provided.

According to the image processing method of the present invention, for each pixel of image data, a noise removal filter having a large filter size is used when the luminance information of the pixel indicates low luminance, and a noise removal filter having a large filter size is used when indicating high luminance. It is characterized in that the noise removal processing is executed using a small noise removal filter.

The image processing method of the present invention can be implemented by a computer, and a computer program therefor is installed or loaded in the computer through various media such as a CD-ROM, a magnetic disk, a semiconductor memory and a communication network. be able to.

[0014]

BEST MODE FOR CARRYING OUT THE INVENTION (First Embodiment) A first embodiment of the present invention will be described with reference to the drawings. FIG. 1 is a block diagram showing the hardware configuration of the printer 1 of this embodiment.

The printer 1 is provided with a power supply mechanism section 2 including a paper feed mechanism 10 for feeding paper into the printer, a print engine 11 for printing, and a paper ejection mechanism 12 for ejecting paper out of the printer. . The print engine 11 includes a paper feeding mechanism, a carriage mechanism, a print head, and the like.

It is the CPU (processor) 13, the ROM 14 and the ROM 14 that control the power mechanism section 2 to perform the printing operation.
RAM15, LCD panel and LCD controller 1
6, PC card slot and PC card controller 1
7, a printer control unit 3 including a communication interface 18 and the like. The CPU 13 uses various means 14 via the internal bus.
To 18 are accessible, and the power mechanism unit 2 is controlled according to the print job sent from the host device via the communication interface 18 to actually perform the printing operation. Note that the power mechanism unit 2 may independently include a CPU, and in that case, the CPU of the power mechanism unit 2
However, it communicates with the CPU 13 via a parallel interface or the like to control the print engine 11 to perform the printing operation.

The structure and operation of the power mechanism unit 2 and the printer control unit 3 are basically the same as the structure and operation of the conventional printer device. However, the printer 1 has a function of performing printing by the printer alone without using the host device. For example, the image compression data such as a JPEG file is read from the external memory through the PC card slot 17 or the like, and the expansion processing is performed. It is configured so that printing can be performed by performing noise removal processing described later or the like (hereinafter, this printing mode is referred to as “direct printing”).

FIG. 2 is a functional block diagram of the printer control unit 3 for realizing direct printing. As shown in the figure, the printer control unit 3 includes a filter size storage unit 20, a panel IF unit 21, and a PC card IF unit 2.
2, a decompression unit 23, a noise removal unit 24, a print data generation unit 25, and the like. Each of the above means is functionally realized by the CPU 13 executing a program stored in the memory.

The filter size storage means 20 stores the correspondence between the brightness information and the size of the noise removal filter. The brightness information is typically a brightness value, for example, 0
It takes a value of ~ 255. However, as will be described later, a G value or the like in the RGB color space can be used as the brightness information. The size of the filter is typically specified by the number of vertical and horizontal elements.

The correspondence relationship between the brightness information and the size of the noise removal filter is determined so that the brightness information indicating lower brightness corresponds to a larger size. An example of the correspondence relationship is shown in FIG. In the example shown in the figure, luminance information indicating low luminance (0 ≦ luminance information Y <16) corresponds to a large filter size (7 × 7 size), and luminance information indicating medium luminance (16 ≦ luminance information Y <32). Corresponds to the next largest filter size (5 × 5 size), and the luminance information (32 ≦ luminance information Y) indicating high luminance is the smaller filter size (3 × 3).
Size).

As can be seen from this example, in the present embodiment, the terms high brightness and low brightness are used as words indicating relative positioning, and "brightness information indicating high brightness (low brightness)". Does not necessarily indicate only high (low) brightness information as an absolute value.

The panel IF means 21 receives commands from the user to the printer 1 via the LCD panel composed of an LCD and operation keys and the LCD controller 16, and outputs the print status and the like to the user. . The command received from the user may be, for example, a command (direct print command) requesting execution of the above-described direct printing.

The PC card IF means 22 is a PC card through the PC card slot and the PC card controller 18.
Data is read / written from / to an external memory (for example, a flash memory such as CompactFlash (registered trademark) or smart media) mounted on the card adapter.

The decompression means 23 decompresses the compressed image data to restore the image data. A well-known process corresponding to the compression method can be used for the decompression process. For example, when the compressed image data is a JPEG file,
For each 8 × 8 block, Huffman code decoding and run-length compression decoding are performed on the AC component, inverse quantization is performed with the DC component to calculate the frequency distribution, and inverse DCT is performed on the frequency distribution. , Restore image data.

The noise removing means 24 performs noise removing processing on the restored image data. Such noise removal processing will be described together with the description of direct printing described later.

The print data generating means 25 subjects the image data to predetermined image processing (for example, color conversion processing, dither processing such as error diffusion method, etc.) which is usually performed in the printing operation, and generates print data.

(Direct Printing) When performing direct printing, it is necessary for the printer 1 to perform noise removal processing while suppressing the processing amount so as not to reduce the printing speed.

Here, since dark current noise or the like is noise added in the direction of increasing the brightness, it has a property that it is not easily noticeable in a bright region, but is easily noticeable in a dark region.

Therefore, in the present invention, the processing amount is suppressed by suppressing the degree of noise removal in a bright area to perform high-speed noise removal, and the degree of noise removal is strengthened in a dark area to perform careful noise removal. The process is configured as follows.

The direct printing process will be described below with reference to the flowchart shown in FIG. It should be noted that each step can be executed by arbitrarily changing the order as long as the processing content does not conflict.

The direct printing starts its processing when the panel IF means 21 receives a direct printing command from the user. At this time, when a plurality of compressed image data are stored in the external memory, for example, those file names are displayed in a list on the LCD panel etc. 16 and the user is allowed to select a desired one.

The PC card IF means 22 refers to the external memory via the PC card slot 18 or the like to obtain the selected image compressed data (S100). When the compressed image data is a JPEG file, it may be partially acquired in block units or block line units.

Next, the decompression means 23 decompresses the acquired image compressed data to restore the image data (or a part thereof) (S101). If uncompressed image data is acquired in S100, the decompression process is omitted.

Next, the noise removing means 24 performs the following noise removing processing on the restored image data (or a part thereof) (S102 to S119).

First, the noise removing means 24 scans the image data to obtain the luminance information Y of the pixel to be processed (S102). Here, the restored image data is YCC
When represented by the color space, the brightness value of the pixel is obtained as the brightness information Y. On the other hand, when the restored image data is represented in the RGB color space, for example, Y =
The brightness information Y is calculated by 0.299R + 0.587G + 0.114B, or the value of the G value is calculated as the brightness information Y. This is because the G value is considered to have a greater correlation with the brightness value than the other components.

Next, the noise removing means 24 determines the luminance information Y.
For low brightness (0 ≦ Y <16) and medium brightness (16 ≦ Y
It is determined which of <32) and high brightness (32 ≦ Y) corresponds (S103).

When the luminance information Y corresponds to high luminance, the noise removing means 24 refers to the filter storing means 20,
A filter size corresponding to brightness information indicating high brightness is selected (S104). In the example shown in FIG. 3, the filter size is 3
× 3 is selected.

Next, the noise removing means 24 obtains the maximum luminance information Ymax and the minimum luminance information Ymin for the image data within the selected size (3 × 3 size) centered on the pixel (S105).

Then, (Ymax-Ymin) is compared with a predetermined threshold value θ (for example, 16) (S106).

When (Ymax-Ymin) <θ, the noise removing means 24 determines that the selected size (3 ×
A filtering process (convolution operation) is performed on the luminance value or the RGB value of the pixel to be processed using a low-pass filter of 3 sizes) (S107). As the low-pass filter, in addition to a simple moving average filter, there are various configurations according to the design, such as a configuration in which the weight of the center of the filter is larger than that of the surroundings, and element values are arranged in a circular shape. A low pass filter can be used.

On the other hand, when (Ymax-Ymin) ≧ θ, the noise removing means 24 uses the median filter of the selected size (3 × 3 size) to determine the brightness value or RGB value of the pixel to be processed. The filtering process is executed on the other hand (S108). This is because when the brightness amplitude within the filter size is large, it is possible to give more texture by removing noise with the median filter.

When the luminance information Y corresponds to medium luminance, the noise removing means 24 refers to the filter storing means 20,
The size of the filter corresponding to the brightness information indicating medium brightness is selected (S109). In the example shown in FIG. 3, the filter size 5 × 5 is selected.

Next, the noise removing means 24 obtains the maximum luminance information Ymax and the minimum luminance information Ymin for the image data within the selected size (5 × 5 size) centered on the pixel (S110).

Then, (Ymax-Ymin) is compared with a predetermined threshold value θ (for example, 16) (S111).

When (Ymax-Ymin) <θ, the noise removing means 24 determines that the selected size (5 ×
Filtering processing is similarly performed using a low-pass filter of size 5 (S112).

On the other hand, when (Ymax-Ymin) ≧ θ, the noise removing means 24 similarly executes the filtering process using the median filter of the selected size (5 × 5 size) (S113).

When the luminance information Y corresponds to low luminance, the noise removing means 24 refers to the filter storing means 20,
The size of the filter corresponding to the brightness information indicating low brightness is selected (S114). In the example shown in FIG. 3, the filter size 7 × 7 is selected.

Next, the noise removing means 24 finds the maximum luminance information Ymax and the minimum luminance information Ymin for the image data within the selected size (7 × 7 size) centered on the pixel (S115).

Then, (Ymax-Ymin) is compared with a predetermined threshold value θ (for example, 16) (S116).

When (Ymax-Ymin) <θ, the noise removing means 24 determines that the selected size (7 ×
Similarly, filtering processing is executed using a low-pass filter of 7 sizes (S117).

On the other hand, when (Ymax-Ymin) ≧ θ, the noise removing means 24 similarly executes the filtering process using the median filter of the selected size (7 × 7 size) (S118).

The noise removing means 24 repeats the above processing until it finishes for each pixel of the image data (S1).
19).

The print data generating means 25 performs predetermined image processing that is usually performed in the printing operation on the image data that has been subjected to the noise removal processing, and generates print data (S120).

Finally, the print data generated in this way is sent to the print engine 11 by the printer control unit 3 and printing is executed in the power mechanism unit 2 (S1).
21).

When the image compression data is partially acquired in S100, S101 to S120 are repeated until the processing is completed for all the image compression data. Further, in this case, the sequentially generated print data may be sequentially sent to the print engine 11, and printing may be executed in parallel with the generation of the print data.

In this way, the filter size is adaptively selected based on the brightness information of each pixel, that is, the large filter size is selected for the low brightness pixel and the small filter size is selected for the high brightness pixel. Since it is configured to perform noise removal processing in a dark area where noise is easily noticeable, noise can be reliably removed based on a large-sized filter to achieve high image quality, while bright noise that is less noticeable In the region, it is possible to perform noise removal at high speed based on a small size filter.
As a result, it is possible to effectively remove noise while suppressing the processing amount of the entire image, and it is possible to execute high-quality printing without unnecessarily reducing the printing speed.

(Others) The present invention is not limited to the above embodiment, but can be variously modified and applied.

For example, the noise removal processing may be configured by fixedly incorporating the correspondence relationship between the brightness information and the filter size. For example, S10 of the direct print processing flow
4, in S109, S114, 3 × in case of high brightness
The processing is configured to fixedly select 3 sizes, 5 × 5 size for medium brightness, and 7 × 7 size for low brightness. In this case, the noise removal processing program itself includes the correspondence relation internally, and the storage means for storing the noise removal processing program becomes the "storage means for storing the correspondence relation between the luminance information and the size of the noise removal filter". It will be applicable.

Further, for example, the filter size storage means 20
With respect to, the filter size and the filter element value may be stored in association with the luminance information. in this case,
It is also conceivable to change the strength of noise removal by changing the weight and arrangement of each element according to the luminance information.

Further, for example, in the above embodiment, the image compressed data is acquired via the PC card slot or the like, but the image compressed data or the uncompressed image is acquired from the host device or a predetermined server via the communication interface. It may be configured to receive data. Further, an electronic device such as a digital camera may be connected to the printer with a cable such as a USB, and the image compression data or the like stored in the electronic device may be directly read.

Further, for example, in the above-described embodiment, the function of printing based on the print job sent from the host device is provided, but it is also possible not to have such a function and only have the direct printing function.

Finally, the image processing technique of the present invention can be applied to devices other than printers, for example, devices such as personal computers and digital cameras. Further, the image forming apparatus of the present invention is not only a device generally called a printer such as an inkjet printer, a laser printer, a label printer, etc., but also various information processing apparatuses (copiers, fax machines, handy terminals) having an image forming function. Etc.) Further, the usage form is not limited to home use, and for example, a usage form is possible in which a printer or the like is installed in a convenience store or the like, and a user brings an external memory or an electronic device to perform printing on the spot. .

[0063]

According to the present invention, since the filter size is adaptively selected on the basis of the luminance information of each pixel to execute the noise removal processing, the processing amount at the time of noise removal can be reduced. Noise can be effectively removed while suppressing, and high-quality printing can be performed without unnecessarily reducing the printing speed.

[Brief description of drawings]

FIG. 1 is a block diagram showing a hardware configuration of a printer according to a first embodiment of the present invention.

FIG. 2 is a block diagram illustrating a functional configuration diagram of a printer control unit according to the first embodiment.

FIG. 3 is a diagram for explaining a correspondence relationship between luminance information and a size of a noise removal filter.

FIG. 4 is a flowchart showing a flow of direct printing processing and noise removal processing.

[Explanation of symbols]

1 printer 2 power mechanism 3 Printer control unit 10 paper feed mechanism 11 Print engine 12 Paper ejection mechanism 13 CPU 14 ROM 15 RAM 16 LCD panel and LCD controller 17 PC card slot and PC card controller 18 Communication interface 20 Filter size storage means 21 panel IF means 22 PC card IF means 23 Extension means 24 Noise removal means 25 Print data generating means

   ─────────────────────────────────────────────────── ─── Continued front page    F-term (reference) 5B057 AA11 BA02 CA08 CA12 CA16                       CB08 CB12 CB16 CC01 CE02                       CE06 CH09                 5C021 PA34 PA38 PA53 PA58 PA78                       RB03 RB08 YA01                 5C077 LL02 PP01 PP02 PP43 PP68                       PQ08 PQ12 TT02

Claims (5)

[Claims] 1. A storage unit for storing a correspondence relationship between brightness information and a size of a noise removal filter, and for each pixel of image data, a size corresponding to the brightness information of the pixel is selected with reference to the correspondence relationship. And a means for executing noise removal processing using the noise removal filter of the selected size. 2. The image processing apparatus according to claim 1, wherein the correspondence relationship is determined such that the luminance information indicating lower luminance corresponds to a larger size. 3. An image forming apparatus comprising the image processing apparatus according to claim 1. 4. For each pixel of image data, a noise removal filter with a large filter size is used when the brightness information of the pixel indicates low brightness, and a noise removal filter with a small filter size is used when the brightness information of the pixel indicates high brightness. And an image processing method for performing noise removal processing. 5. A program for causing a computer to execute the image processing method according to claim 4.