JP3935648B2 - Isolated point remover - Google Patents

Description

[0001]
BACKGROUND OF THE INVENTION
The present invention relates to an isolated point removal apparatus that removes isolated points by a SIMD (Single Instruction Multiple Data Stream) control method that processes a large number of pixels at once.
[0002]
[Prior art]
In general, an isolated point is generated on a background portion of an original due to toner scattering or the like. As a conventional example for detecting this isolated point, for example, as shown in Japanese Patent Laid-Open No. 6-6917, a method of detecting an isolated point based on the run length of a white pixel section and a black pixel section (hereinafter referred to as a run length method) As disclosed in Japanese Patent Application Laid-Open No. 8-305844, a method for detecting an isolated point based on density data of pixels constituting an area including a target pixel (hereinafter referred to as area method) has been proposed. Specifically, in this area method, four types of areas including a pixel of interest at the center are formed in a 5 × 5 pixel matrix, and pixel data in the area are compared with the pixel of interest.
[0003]
Here, these actual circuits are configured by an ASIC that can process image data in which data for one pixel is input every pixel clock. FIG. 9 shows an image processing unit 202 of a conventional copying machine that processes image data in which data for one pixel is input every one pixel clock. The filter unit 203 and the isolated point removing unit 204 perform filter calculation and isolated point removal in parallel on the image data read by the scanner 201, and select each output by the selector 205. The image data selected by the selector 205 is output to a printer (not shown) via the gamma processing unit 206 and the gradation processing unit 207. Here, although the filter unit 203 is omitted in the drawing, an enhancement processing unit that performs enhancement processing on image data, a smoothing processing unit that performs smoothing processing, and a character region by determining whether or not it is a character region In this case, the output of the enhancement processing unit is selected, and when it is not a character area, the output of the smoothing processing unit is selected and output to the selector 205.
[0004]
[Problems to be solved by the invention]
The former “run-length method” requires less memory capacity and is suitable for sequential processing for each dot, but is effective in the SIMD control method that simultaneously processes the same instruction for a fixed amount of image data. There is a problem that there are few. Further, the latter “area method” forms four types of areas including the target pixel and compares the pixel data in the area with the target pixel, so that the SIMD control method is difficult to process. is there. That is, the conventional method as described above is considered on the premise that data of one pixel is input every one pixel clock, and thus is difficult to apply to the SIMD control method.
[0005]
In view of the above-described problems of the conventional example, the present invention eliminates isolated points by an SIMD control method in which data processing of several hundred dots is performed by one instruction instead of data processing for each dot. An object is to provide an apparatus.
[0006]
Another object of the present invention is to provide an isolated point removal apparatus that can increase the processing speed by reducing the number of instructions.
[0007]
Another object of the present invention is to provide an isolated point removing apparatus capable of adjusting the degree of removal of isolated points.
[0009]
[Means for Solving the Problems]
In order to achieve the above object, the first means is such that the density of the outer peripheral pixels in the matrix of m × n (m ≧ n) pixels centered on the target pixel is less than or less than the first threshold, and It is determined whether or not the density of a plurality of pixels on the left and right sides of the same line as the pixel of interest outside the matrix is less than or less than a second threshold value. It has a means to determine.
[0010]
In order to achieve the above object, the second means is such that the density of the outer peripheral pixels in the matrix of m × n (m ≧ n) pixels with the target pixel as the central pixel is all equal to or less than the first threshold value, and It is determined whether or not the density of every other pixel on the left and right of the same line as the pixel of interest outside the matrix is less than or less than a second threshold, and It has a means to determine a pixel as an isolated point.
The third means is characterized in that the first and second threshold values are set to be variable in the first and second means.
[0011]
DETAILED DESCRIPTION OF THE INVENTION
Embodiments of the present invention will be described below with reference to the drawings. FIG. 1 is a block diagram showing an embodiment of an isolated point removing apparatus according to the present invention, FIG. 2 is an explanatory diagram showing a SIMD control system, FIG. 3 is an explanatory diagram showing reference pixels for isolated point determination, and FIG. FIG. 5 is an explanatory diagram showing an isolated point determination flag.
[0012]
In FIG. 1, the filter 103 performs enhancement processing and smoothing processing on the image data read by the scanner 101, and determines whether or not it is a character region and performs enhancement processing in the case of a character region. Image data is selected. If it is not a character area, smoothed image data is selected. The isolated point removing unit 104 determines whether the image data read by the scanner 101 is an isolated point, accumulates a plurality of isolated point determination results, and sets the processing result of the filter 103 to “0” in the case of an isolated point. Select data that has been converted to or smoothed. The output signal of the filter 103 is output to a printer (not shown) via the gamma processing unit 105 and the gradation processing unit 106.
[0013]
Here, the conventional image processing unit 202 shown in FIG. 9 inputs / outputs image data based on the clock, and is configured mainly with an ASIC operating as hardware. However, the SIMD control method shown in FIG. In the case of the image processing unit 102, an instruction indicating processing contents can be described by software. FIG. 2 shows processing of the SIMD control method. First, input image data from the scanner 101 is loaded (301), and then a certain amount (k dots) of data that can be processed in batch by the SIMD control method is loaded (302). The instruction is executed every time, and the processed data is output by one dot d (303).
[0014]
Here, in the conventional “run-length method”, in order to detect white length (continuous amount of white pixels) and black length (continuous amount of black pixels), data of one pixel input every clock is used as a threshold value. If the white pixel counter and the black pixel counter are used for comparison, no memory is required. However, when one instruction is executed for a certain amount of pixel data as shown in FIG. 1, the counter method is applied. I can't. In contrast, the SIMD control method can process a large number of image data at a time. Here, since it is more effective to reduce the instruction “compare with threshold”, the present invention focuses only on white pixels.
[0015]
Next, the isolated point removal process will be described in detail with reference to FIGS. In this example, in FIG. 3, a surrounding pixel 401 indicated by a circle with respect to a 5 × 5 pixel matrix including the target pixel P, and the number of the same lines as the target pixel P outside the matrix as indicated by a × mark. Reference is made to the pixels 402L and 402R. Then, as shown in FIG. 4, when the values of all surrounding pixels 401 are smaller than the threshold value TH1 (or below the threshold value TH1), it is determined that the surrounding pixels 401 indicated by a circle are not halftone dot images or photographic image areas. At the same time (steps S1 to S3), if the values of all the pixels 402L and 402R indicated by x are smaller than the threshold value TH2 (or less than or equal to the threshold value TH2), there are no characters or halftone dots around the target pixel P. A region is determined (steps S4 to S6).
[0016]
As shown in FIG. 5, this determination result is represented by a 2-bit flag for each pixel, and represents the comparison result between the surrounding pixel 401 indicated by a circle by the first bit (MSB) and the threshold value TH1, and 2 bits. A comparison result between the pixels 402L and 402R indicated by x and the threshold value TH2 is represented by the eye (LSB). When these two conditions are satisfied, it is determined that the region where the target pixel P exists is a background portion, and isolated points are removed (steps S7 to S9).
[0017]
Next, the isolated point removal processing of the second embodiment will be described in detail with reference to FIG. In this example, first, with respect to a 5 × 5 pixel matrix including the target pixel P in FIG. 6, the values of all surrounding pixels 401 indicated by ◯ are smaller than the threshold value TH1 (or less than the threshold value TH1). As in the first embodiment, it is determined that the surrounding pixel 401 is not a halftone image or photographic image area.
[0018]
In the second embodiment, the values of every few pixels 502L and 502R outside the matrix and on the same line as the pixel of interest P are compared with the threshold value TH2 as shown by x in FIG. If every other pixel 502L, 502R is smaller than the threshold value TH2 (or less than the threshold value TH2), it is determined that there is no character or halftone dot around the pixel of interest P. The reason is that pixel data smaller than the threshold value TH2 (or less than the threshold value TH2) are arranged every other pixel on the assumption that there is no halftone dot portion, photographic portion, or character portion where black and white are switched for each pixel. This is because it is determined that there is no character or halftone dot around the pixel of interest P. According to the second embodiment, the number of computations for comparison with the threshold value TH2 can be reduced as compared with the first embodiment, so that the processing time can be shortened.
[0019]
Next, the isolated point removal process of the third embodiment will be described in detail with reference to FIGS. Here, since the background density of the document is not always “0”, the threshold TH2 needs to be a value at which the background portion can be determined as an area where white pixels are continuous even at the lowest. Also. Since the threshold TH1 refers to the surrounding pixels 401 of the pixel of interest P, if the value is smaller than the background density, an isolated point cannot be determined. Therefore, at least the threshold TH1 needs to be the same value as the threshold TH2. And by making these two threshold values TH1 and TH2 values that vary independently, as shown in FIG. 8, the larger the threshold values TH1 and TH2, the easier it is to remove noise. The smaller the thresholds TH1 and TH2, the more difficult it is to remove noise.
[0020]
Here, the “same line” in the above embodiment refers to a line in the main scanning direction. The main scanning direction is the electrical reading direction of the scanner (CCD longitudinal direction), and the sub-scanning direction is the mechanical reading direction of the scanner (CCD moving direction). The matrix circle of the present invention is not limited to the quadrangular shape as in the above embodiment, and may be a substantially circular shape (octagonal shape) obtained by cutting off the corners of the quadrangular shape.
[0022]
【The invention's effect】
As described above , according to the first aspect of the present invention, the density of the outer peripheral pixels in the matrix of m × n (m ≧ n) pixels having the target pixel as the central pixel is all equal to or less than the first threshold value. And whether or not the density of a plurality of pixels on the left and right sides of the same line as the target pixel is outside or below the second threshold is less than or less than the second threshold value. Since it is determined as an isolated point, the isolated point can be removed by the SIMD control method.
[0023]
According to the second aspect of the present invention, the densities of the outer peripheral pixels in the matrix of m × n (m ≧ n) pixels centered on the target pixel are all equal to or lower than the first threshold value, and It is determined whether or not the density of every other pixel on the left and right of the same line as the target pixel is all equal to or less than the second threshold value, and if the condition is satisfied, the target pixel is determined to be an isolated point Therefore, the isolated points can be removed by the SIMD control method, and the processing speed can be increased by reducing the number of instructions.
[0024]
According to the third aspect of the invention, since the first and second threshold values can be changed, it is possible to remove isolated points by the SIMD control method, and to adjust the degree of removal of isolated points. it can.
[Brief description of the drawings]
FIG. 1 is a block diagram showing an embodiment of an isolated point removing apparatus according to the present invention.
FIG. 2 is an explanatory diagram showing a SIMD control method.
FIG. 3 is an explanatory diagram showing reference pixels for isolated point determination.
FIG. 4 is a flowchart showing isolated point removal.
FIG. 5 is an explanatory diagram showing an isolated point determination flag.
FIG. 6 is an explanatory diagram showing a reference pixel for isolated point determination in the second embodiment.
FIG. 7 is an explanatory diagram showing a threshold for isolated point determination in the third embodiment.
FIG. 8 is an explanatory diagram showing a relationship between a threshold value and how isolated points are removed.
FIG. 9 is a block diagram showing a conventional isolated point removing apparatus.
[Explanation of symbols]
102 SIMD image processing unit 103 Filter 104 Isolated point removal unit 105 γ processing unit 106 Gradation processing unit

Claims (3)

The density of all the peripheral pixels in the matrix of m × n (m ≧ n) pixels centered on the target pixel is less than or less than the first threshold value and is outside the matrix and on the same line as the target pixel. An isolated point removing apparatus comprising: means for determining whether or not the densities of a plurality of left and right pixels are all equal to or less than a second threshold and determining the target pixel as an isolated point when the condition is satisfied . The density of all the peripheral pixels in the matrix of m × n (m ≧ n) pixels centered on the target pixel is less than or less than the first threshold, and is outside the matrix and on the same line as the target pixel. An isolated point comprising means for determining whether or not the density of every other plurality of pixels on the left and right is less than or less than a second threshold and determining the pixel of interest as an isolated point when the condition is satisfied Removal device. 4. The isolated point removing apparatus according to claim 2, wherein the first and second threshold values are variable .

Images