MXPA98008694A - Low-resolution masking adjustable dynamically for the processing of digital images - Google Patents

Abstract

This invention relates to the processing and digital images and, more particularly, to a dynamically adjustable technique, to reduce the force of and blur the images.

Description

LOW RESOLUTION MASKING ADJUSTABLE DYNAMICALLY FOR THE PROCESSING OF DIGITAL IMAGES Field of the Invention This invention relates to the processing of digital images and, more particularly, to a technique, dynamically adjustable, to reduce the force of and blur the images.

Cross Reference to the Related Request This application relates to a United States patent application filed at the same time as the present one, commonly granted by Donald J. Curry, No. of Series on "Digital Image Formation with Edge Sharpening for a Print Mode that Saves Organic Pigment Ink".

Background and Brief Description of the Invention Low-resolution masking is used in photography for edges of intensely illuminated images. Up to this point, a blurred or unfocused version of an image is subtracted from the original image to produce a model of the low resolution image. This mask is then superimposed on the original image to REF: 28296 increase the contrast of any abrupt contrast transitions (ie, edges) within the image. This procedure can be described by a linear equation: I exit = lent. "•" (l ent ca? ~ nae) ra, where ITI-, = original image image = blurred image C = a variable sharpening / blur control Isalida = sharpened edge image Digital images are composed of discrete tones (for example, often only black and white pixels), so the perceived quality of those images depends to a large extent on the human visual response to the printed pixel patterns that define the images. In this way, continuous-tone images are more effectively represented by pixel patterns, such as mid-tone pixel patterns, which are optimized to create the appearance of uniformity or smoothness, well-controlled contrast variations, while linear line images, such as texts, are best represented by pixel patterns that are optimized to create the appearance of fine contrast and resolution variations. Therefore, it would be beneficial to be able to dynamically tune the mask of a low resolution digital image for the type of digital image being printed, so that intense illumination of the edge of the low resolution mask can be obtained without counteracting the perceived quality of the digitally printed image as a whole. • Sometimes the quality of the perceived image is secondary to the goal of printing the fundamental information of the contra < It's the image at the lowest possible cost. Some printers have a "draft" print mode for this purpose. The fundamental information and contrast of many images is represented by the edges of the elements that make up that image. Consequently, it would be beneficial to have a technique to generate effectively "differentiated" versions of the digital images, so that marking material (eg, organic pigment or ink) can be saved when printing those images in a "draft" mode.

BRIEF DESCRIPTION OF THE DRAWINGS Other objects and advantages of this invention will become apparent when the following detailed description is read in conjunction with the accompanying drawings, in which Figure 1 is a block diagram of a digital image processing system with masking of low tuneable resolution according to this invention; Figure 2 is a block diagram of a stage representative of a separation filter of the type that can be used to reduce the computational complexity to carry out this invention; and Figure 3 is a tone reproduction curve for an organic pigment saving mode of operation.

Detailed Description of the Invention Although the invention is described in some detail herein with reference to a specific embodiment, it should be understood that it is not intended to limit this to that modality. On the contrary, the interest is to cover all modifications, alternatives and equivalents that fall within the spirit and scope of the invention as defined by the appended claims. Returning now to the drawings, and at this point especially to Figure 1, it depicts an original image at an appropriate resolution, such as 300 dots per inch (dpi) per pixel values (typically 8 bit values) that they are received in series in order of frame scanning (ie, order from left to right / top to bottom) of an appropriate image feed terminal (IIT). As a general rule, the IIT is a scanning subsystem. In order to carry out the low resolution masking of the original images that are contemplated by this invention, there is a resampling of images, the contextually sensitive flattening filter 21 to feed a blurred or unfocused version of the original image to the generator of a masking mask. low resolution 22. In the generator of the low resolution mask 22, a vision device 23 calculates the difference between the original image and the blurred or unfocused version of the image to provide a low resolution mask of the original image. A variable multiplier 24 then weights this low resolution mask according to an adjustable weighting factor F. Finally, an addition device 25 then combines, additively, this low resolution mask weighted with the original image to transform the image low-resolution masked at the selected magnification and resolution by means of a suitable transformation machine 26. As illustrated, a SRAM memory 28 functions as a randomly addressable FIFO angular absorber, to share the pixel values that are necessary to calculate the Blurred or unfocused and low resolution masked versions of the original image. In one application, it has been found that only about five context scan lines are required to satisfactorily calculate the blurred or unfocused image, so that the SRAM 28 is dimensioned accordingly. The data is transferred back and forth between the SRAM 28 and an SRAM 29 interconnect via the SD data path, and the places within which the reading and writing of that data are performed are determined by the addresses contained in a address path SA. To reduce computational complexity, the flattening filter 21 is a so-called separation filter having two sections, 31 and 32, one to contextually flatten the image in the slow scan direction (ie, the processing direction) and the other to contextually flatten the image in the orthogonal or fast scan direction. Thus, as shown, the original fed image is first re-sampled by a linear interpolator 34 and then flattened in the slow scan direction by the first section 31 of the filter 21. This provides an intermediate representation of the image, so that the sufficient context of this intermediate representation to carry out the flattening of the image in the fast scan direction is stored in a shift register device 35 which has an output after each of its stages. In this way, the fast scan context that is accumulated by the shift register device 35 is resampled by another linear interpolator 36 and then flattened in the fast scan direction by the second layer 32 of the smoothing filter 21. Typically, each The filter section 21 comprises, as shown in Figure 2, the multipliers 61-65 to multiply the current value of the pixel and its closest vertical or horizontal neighbors by the weights or values of 2., 2, 2, lyl respectively, (or 1,2,2,2,1 as seen with the filter centered on the current pixel). Those weighted values of the pixel are then added by the addition device 67-70. Finally, the aggregate of the sums is divided by the sum of the weights or values in a divisor 71. As is known, multiplying and dividing powers of two corresponds to displacements in one direction or another when they are implemented in the equipment. As will be appreciated, the original image is effectively graded through the smoothing filter 21 in the context and at a desired magnification and resolution. The amplification can cause the image to be graded through the smoothing filter 21 in steps that are not integer multiples of the original scan step, so that a shift compensator 38 is provided to make the linear resampling interpolators 34 and 36 substantially compensate for any displacement of the positions of the center of the pixel in the amplified image of the central positions of the corresponding pixels in the original image. An additional context line is usually sufficient to allow interpolator 34 to perform this compensation in the slow scan direction. In the slow scan direction, however, the interpolator 36 may not only have to compensate for the amplification caused by deviations from the center of the pixel, but may also have to compensate for resolution misalignments (such as when an image is scanned at 300 dpi, is transformed to 600 dpi). An image processing line is used to carry out this invention, so that the data is transferred synchronously from one step of the process to another. The recording devices 41 and 42 through which the blurred or unfocused image focused and the original image are respectively displaced to calculate the low resolution mask and the low resolution masked output image that are part of this line. According to the present invention, the weighting factor F for the low resolution mask can be adjusted by varying the multiplication of the multiplier 24. This weighting can be adjusted dynamically on, say, a pixel base per pixel in response to, say, a signal characterization of the type of a logical image supplied by an image segmenter (not shown). Alternatively, the weighting factor F of the multiplier 24 can be controlled in response to the manual manipulation of a mode switch (it is not shown either). 0, the dynamic and manual control of the weighting factor F can be combined. The low resolution mask tends to soften or sharpen the transformed image depending on whether the weighting factor F is negative or positive, respectively. Thus, for example, the weighting factor F can be selected such that it is approximately (a) -1.0 to suppress and / or flatten half-tone representations of continuous-tone images, (b) 1.2 for printing of linear stroke images to increase the edges of the image, or (c) 2.0 or greater for printing in draft mode to make the edge resolution strongly dominate the transformation in the interest of saving marking material. If the weighting factor F is controllable only manually, it can be set to a value of approximately 1.0 for a normal printing mode and a value of approximately 2.0 for a draft mode or organic pigment / ink saving. As shown, the values of the output pixel are adjusted to provide the desired transformation function for each selected operating mode of the transformation machines by a mode that depends on the look-up table 60. For example, as shown in FIG. Figure 3, there is an extreme example of the so-called tone reproduction curve that is characteristic of the operation of the organic pigment saving mode. It is noted that in relation to this date, the best method known by the applicant to carry out the aforementioned invention, is that which is clear from the present description of the invention. Having described the invention as above, property is claimed as contained in the following:

Claims (1)

1. A dynamically adjustable low resolution masking procedure for perceptually altering the edges of images, of digitally transformed images, depending on the distinguishable characteristics of the images, the procedure is characterized in that it comprises: generating a blurred or unfocused representation digitally represented in an image original digitally represented; compare the digital representation of the original image with the digital representation of the blurry or unfocused version of the original image to define a low resolution digital mask for such an image; multiply the low resolution mask by an adjustable weighting factor, the weighting factor can be adjusted dynamically depending on the characteristics of the image to blur or blur the edges in some regions of the image and emphasize the edges in other regions of the image; and combine the low resolution mask weighted with the original image to transform a low resolution masked version of the original image.