X-ray or gamma ray systems or devices – Electronic circuit – With display or signaling
Reexamination Certificate
1999-10-22
2001-06-26
Kim, Robert H. (Department: 2882)
X-ray or gamma ray systems or devices
Electronic circuit
With display or signaling
C382S274000
Reexamination Certificate
active
06252931
ABSTRACT:
BACKGROUND OF THE INVENTION
1. Field of the Invention
The invention relates to a method for processing an original image, which method includes the following steps: decomposition of the original image into a series of detail images, each of which contains only a part of the spatial frequencies of the original image, modification of at least a part of the image values associated with the pixels of at least one of the detail images, reconstruction of a final image from the possibly filtered detail images.
2. Description of Related Art
The original image may be an X-ray image, but it is also possible to process other images produced for medical diagnostics and available an electronic form, in which an image value which characterizes the brightness in the relevant pixel is associated with each pixel (a pixel is an image element; an image is composed of a matrix of pixels). The invention also relates to a device for carrying out such a method.
A method of the kind set forth is known from EP-A 527 525. According to the known method the original image is subjected to low-pass filtering at a first decomposition level and the low-pass image thus formed is subtracted from the original image, thus yielding a difference image which is referred to as a first detail image in which only the very highest spatial frequencies of the original image are preserved. Because of the subtraction of the low-pass image from the original image, positive as well as negative values may occur for pixels in the detail image. These values represent the contrast of the image; however, hereinafter they are also referred to as image values.
At a second decomposition level the low-pass image is subjected to a further low-pass filtering operation, thus producing a second low-pass image which, in relation to the original image, represents spatial frequency components which are less high than those represented by the first low-pass image. The second low-pass image is subtracted from the first low-pass image, thus yielding a second detail image which represents mainly spatial frequencies which lie below the spatial frequency band of the first detail image but above the spatial frequency range of the second low-pass image.
At a third decomposition level a third low-pass image containing even less high frequency spatial frequency components is formed from the second low-pass image in an analogous manner and a third detail image is derived from the difference between the second and the third low-pass image. The spatial frequency band represented by such a third detail image lies mainly below the spatial frequency range of the second detail image but above the spatial frequency range of the third low-pass image. The third decomposition level is succeeded by further decomposition levels, the detail images then produced representing each time spatial frequencies which are lower than those of the detail image of the preceding decomposition level.
The detail images and the residual image are added after at least one of the detail images has been modified. The aim is to enhance the diagnostically relevant image information and to suppress the image information which is not important for the diagnosis or has a disturbing effect.
EP-A 527 525 describes two alternatives for the modification of the detail images:
a) According to a first alternative, the image values of a detail image are multiplied, one pixel after the other, by a factor which is dependent on an image value of a pixel having corresponding co-ordinates in a low-pass image whose spatial frequency range is below the spatial frequency band of the detail image. For small image values in this low-pass image (corresponding to bright image areas in the case of an X-ray image) the factor derived from a look-up table is smaller than 1, whereas it is larger than 1 for large image values. This reduces the contrast in the bright image areas where noise is particularly disturbing whereas the contrast in the darker image areas is enhanced. However, the reduction and the enhancement are carried out independently of the image values in the detail image. The noise in these image areas, therefore, is reduced in the same manner as the useful information contained in the larger image values.
b) According to the second alternative the image values of at least one detail image are modified in dependence on their magnitude, in conformity with a non-linear characteristic, so that the very small image values (corresponding to the noise) are attenuated, moderately small image values are enhanced and larger image values remain the same. However, this modification is performed irrespective of the fact whether the image value is associated with a pixel which lies in the bright or in the dark image areas. In the darker image areas, in which the effect of noise is less than in the bright image areas, small image values of the detail image are thus attenuated in relation to the larger image values. This method has the drawbacks that it is not robust in relation to noise pulses which no longer lie in the range of the very low contrasts but in the range of moderately low contrasts and hence are even intensified, whereas useful information in the range of very low contrasts is attenuated in any case.
Citation of a reference herein, or throughout this specification, is not to construed as an admission that such reference is prior art to the Applicant's invention of the invention subsequently claimed.
SUMMARY OF THE INVENTION
It is an object of the present invention to provide a method of the kind set forth whereby the image quality can be enhanced even further. This object is achieved according to the invention in that the modification of an image value corresponds to multiplication of this image value by a factor which is dependent on this image value itself as well as on the mean image value and/or on the variation of the image values in a window around the relevant pixel in the original image or an image derived therefrom.
The invention is based on the recognition of the fact that optimum image quality can be achieved only if the modification of the image values is dependent on the relevant image value itself as well as on the mean image value and/or the variation of the image values in a window around the relevant pixel in the original image or an image derived therefrom. This enables, for example, the small image values in a detail image to be intensified in relation to the larger image values when they are associated with a pixel which is situated in a dark image area, whereas they are not intensified when the relevant pixel is situated in a bright image area.
The invention is also based on the recognition of the fact that instead of being dependent on the mean image value (being the same as the density of the image) in a window around the relevant pixel, the modification may also be dependent on the variation of the image values in such a window; this is notably so when the aim is to avoid enhancement of the noise upon enhancement of the contrast. This is because in image areas in which the image values vary strongly the noise is not observed as clearly as in image areas with only a slight variation of the image values.
One possibility of implementing the invention would be to provide a look-up table in which a factor is stored for each image value of a pixel and for each density (mean image value) in a window around this pixel. Taking up such a dual dependency in a look-up table, however, would be very intricate and modifications of such a dual dependency could be implemented only at great expenditure.
Therefore, in an essentially simpler possibility, the modification includes converting the image values in at least one of the detail images by means of a non-linear characteristic in such a manner that the small image values in the detail image are increased in relation to the larger image values, and controlling the degree of increase of the small image values in the pixels of the detail image in dependence on the mean image value and/or the variation of the image values in a window ar
Aach Til
Buzug Thorsten
Maack Ingo
Pralow Thomas
Springub Ansgar
Kiknadze Irakli
Kim Robert H.
U.S. Philips Corporation
Vodopia John F.
LandOfFree
Processing method for an original image does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Processing method for an original image, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Processing method for an original image will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-2529179