DE60308282T2 - Image processing device - Google Patents

Description

object the invention

The The present invention relates to an image processing apparatus and methods for improving color images.

background the invention

For the most Applications requires the amount of color images to render required data that some form of compression or the bandwidth reduction is done to reduce the amount of data which transmit or must be stored. That's because of the color images at least a representation of the red, green and blue (RGB) components to be transferred or image to be saved to make a truthful To ensure representation of the original colors of the picture. The next generation mobile wireless phones may require the ability to Color pictures over a wireless connection for viewing on a as part of the phone executed Receive and display video. It is necessary in a similar way in applications such as personal Digital Assistant (PDA) to transfer color images for the screen and save. For such applications can the color images are either received in compressed form or compressed after receiving to reduce the amount of data the one to transfer and saving the image is required.

A Technique for reducing the amount of information needed to store a representation of a color image is required, is the dynamic bandwidth the RGB signal samples in which a number of those used for the representation of RGB components Bits is reduced. For example, if each pixel is shown as 18-bit, is for Each RGB component is assigned a 6-bit value, which is an area perimeter of 64 possible color values for every Component, producing 262,144 different colors are. If the number of used to represent each component Bits are reduced to 4-bit, each pixel can be represented with 12-bit which results in a 33% reduction in saving the color image required memory size implemented becomes. However, as a result of this reduction, the number is potential Colors that can be displayed per pixel are reduced to 4096, which means the so-called "color palette" available for displaying the image is reduced. As a result, the quality of the stored image during reproduction are affected. For example, can certain pictures, such as faces or sunsets below Contour effects that result in certain parts of the image the restriction the available Colors appear that are due to the reduced color palette. Preferably, a reduction should be used to represent the signal samples used number of bits without reducing the quality of the reproduced Picture.

"Bit depth" (Stephen Canning) is an article that describes how it is possible to save a needed digital image To reduce the amount of data by reducing its bit field depth. The concept of bit depth is explained, a list that shows like the change bitfield depth changes the number of colors that can be displayed per pixel specified, and it is emphasized that a change in bit depth by Storage of an image as a GIF file can be achieved.

Summary the invention

The The present invention is defined by the appended claims.

According to the present invention, there is provided an image processing apparatus operable to process input color component signal samples representing an input color image to produce output color component signal samples representing an enhanced version of the input color image. The apparatus includes an expansion processor operable to extend the dynamic range of the color component signal samples, a low-pass filter ( 36 ) operable to filter the extended dynamic range color signal samples to form the output color component signal samples, and a detector processor. The detector processor is operable to detect color component signal samples associated with a detail representation appearing in the input-side color image. According to the detection of the detail in the image, the detector processor is operable to reduce the bandwidth of the low-pass filter (FIG. 36 ) is adjustable in accordance with the detected detail.

Embodiments of the present invention are designed to provide a way of reducing the amount of data needed to represent a color image, thereby increasing or at least preserving the quality of the image during playback. Finally, an image processor embodying the present invention is equipped with a color enhancement processor arranged to expand the dynamic range at which the color components of the image are represented by increasing the number of bits used to represent the RGB color component samples are used. The expanded in their dynamic range samples are then with filtered out a low-pass filter to ensure a smooth transition between the colors of the reproduced image. Nevertheless, to reduce the likelihood of loss of detail within the reproduced image, a detector processor is arranged to detect signal samples representing the detail within the image and to adjust the bandwidth of the low pass filter according to the detected detail. In one embodiment, the samples representing the detail are arranged to bypass the low-pass filter. As a result, the likelihood of losing detail in the image is reduced, thereby also reducing the likelihood of contour effects that can occur in the image due to a reduced color gamut caused by a limited dynamic range of the color component signal sample.

Further advantageous embodiments The present invention will be described by describing the embodiments vividly. In particular, the image processor that processes the color images with an extended dynamic range, with a small number of components are running, making a cost-effective execution is implemented.

Summary the drawings

versions The present invention will now be described by way of example only with reference to the associated Drawings are described, wherein like parts with corresponding Reference numerals are designated, in which:

1 a block diagram of an LCD screen module is;

2 Fig. 12 is a block diagram of an LCD screen module including an image processor according to an embodiment of the invention;

3 is an illustration of a color image example suffering from contour effects resulting from a limited color palette;

4 a schematic block diagram of the in 2 pictured image processor;

5 a schematic block diagram of the in 4 mapped dynamic range converter;

6 is a table representing values required to supplement the signal samples corresponding to the expanded dynamic range in consideration of the corresponding expansion produced by copying the high-order bits into the low-order bits; '

7 is a schematic representation of an impulse response of a moving average filter of eight taps duration;

8th a schematic block diagram of the in 4 pictured low pass filter;

9 a schematic block diagram of the in 4 imaged detector processor;

10 is a flow chart illustrating the operation of the in 4 represented image processor represented; and

11 a representation of the color image example after processing by the in 3 shown image processor is.

Description of the preferred versions

An embodiment of the present invention will now be described with reference to an LCD screen module. 1 provides an illustration of a typical LCD screen module, one that would be used in a mobile, wireless telephone or personal digital assistant. As the mobile wireless phones move towards the next generation, the so-called third generation (3G), there will be a growing need to display digital images on a LCD screen module of mobile wireless phones. A typical screen module for this application is an LCD display panel and is in 1 such a module comprises a display memory 1 which is adapted to store signal samples associated with a digital image displayed on the panel of the LCD display 2 should be displayed. Pixels associated with the color image to be displayed are on a connection channel 4 received and in the display memory 1 by means of the write address logic 8th written. If the image is to be displayed, they are in memory 1 stored signal samples by read address logic 10 read out and into analog values through a digital to analogue converter 12 before on the LCD panel 2 to be displayed, converted.

As already explained, the present invention offers a possibility for reducing the amount of data to be stored, that is to say the size of the display memory 1 in the present example, while the quality of, for example, on the LCD panel 2 image to be displayed. The reduction of the amount of data to be stored is achieved by reducing the dynamic range of the signal samples associated with the color components. Alternatively, the image may be received with component signal samples containing a due to shortening have reduced dynamic range, which may be required to transfer the image. As will be appreciated, the RGB components are shortened by the same amount. In both of these examples, an image processor embodying the present invention is arranged to enhance a representation of the received or stored color image. As explained, the enhancement may promote a true or at least improved reproduction of the original image with a reduced likelihood of the contour effects that may occur in the image.

An embodiment of the present invention is in 2 illustrated. 2 shows the LCD display module 1 in combination with an image processing circuit 20 , As shown in 2 , are the pixels of the color image, as on the channel 4 received, preferably by preferably 12 bits instead of 18 bits, as shown in 1 , represented. The reduction of the resolution of the signal samples can be achieved by the color reduction processor 15 be executed. The color reduction processor 15 can restrict the dynamic range of the signal sample by shortening the signal sample from 18 bits to 12 bits by discarding the last two least significant bits of each RGB component. As such, the RGB color components are advantageously represented as 4 bits instead of 6 bits, resulting in a 33% reduction in the memory capacity requirement of the display memory 1 results.

A result of reducing the dynamic range of the color signal samples by truncating 6 bits to 4 bits is the reduced color palette available to display the re-rendered image. As a result, a color image will suffer quality losses, which may cause the appearance of contours or apparent flat sectors in the reproduced image. An example of such a color image suffering from contour effects due to reduced color gamut is in FIG 3 shown.

In order to reduce the likelihood of contour effects occurring in a reproduced image, the display module of the present invention is an image processor 20 fitted. The image processor 20 is set up to control the dynamic range of color component signal samples coming from the display memory 1 be read out, expand. As in 2 For the illustrated embodiment, the number of bits used to represent each pixel is increased from 4 bits to 6 bits. The image processor 20 is more detailed in 4 shown.

In 4 the image processor receives 20 the 4-bit RGB color component signal samples on the connection channel 22 the read address logic 10 , The color component signal samples are to the color converter 30 for the dynamic range and a detector processor 34 fed in parallel. The color converter 30 for the dynamic range is arranged to increase the number of bits for representing color component signal samples from 4 bits to 6 bits.

The extended color component signal samples are then a low pass filter 36 after which the extended color component signal samples are sent to an output channel 24 are fed. As shown in 2 , the extended signal samples are a digital to analog converter 12 fed into the 4 is shown.

The detector processor 34 is arranged to generate a filter selection control signal indicative of a channel 38 the low-pass filter 36 is supplied.

In accordance with the color converter 30 For the dynamic range generated expanded dynamic range, the values of the 4-bit signal samples can be scaled to reflect the expanded dynamic range that results from increasing the number of bits to six. In order to adjust the color component signal samples according to the extended dynamic range, the values of the input signal samples are multiplied by a gain factor G. The gain factor G is determined according to the original resolution of n-bit and the increase of the resolution by q-bit according to the following formula:

To reduce the number of components required to determine the gain factor G, the gain factor G is generated by copying the two highest quality bits from the input 4-bit sample into the two least significant bits of the 6-bit output sample. Such a means for adding a gain value to the input side sample is disclosed in US Pat 5 illustrated. In 5 For example, the input 4-bit sample is assigned the letter ABCD in a four-level register 40 are stored. A six-step register 42 is provided for forming an output signal sample. The connection channels 44 connect the quadruple register to the sixfold register. As shown in 5 are the connection channels for the 2-bit values A, B in the first two stages of the 1 shift register, which are in the last two stages of the register 42 have to be copied. Thus, the values of the two least significant bits of the output signal sample are formed from the highest quality bits of the input side signal sample.

6 provides a table that allows all shows the values of the input 4-bit signal samples containing gain values required to adjust 4-bit to 6-bit signal samples to take advantage of dynamic range expansion. The gain value as calculated by the above formula is shown depending on the gain value provided by the bit map corresponding to the means of the in 5 generated register is generated. As in 6 can be seen, the bit mapper makes a good approximation to the calculated gain factor G, but has the advantage of not requiring multiplier components, thereby reducing the cost of execution.

The extended color component signal samples are through a low pass filter 36 filtered to ensure a smooth transition between the color values. The filtering of the signal sample works has the effect of reducing the likelihood of contour effects that can occur in the rendered image. In advantageous embodiments, the low-pass filter 36 is implemented as a moving average filter having an impulse response as in 7 shown. 7 illustrates an impulse response from a moving eight tap averaging filter with an effective duration T _w .

One moving average filter effectively shapes an output side Signal sample from the mean values of the input signal samples within one by the impulse response of the moving average filter shaped window. The filter is set up a smoothing Effect on the contours of the image due to the restricted dynamic Range of resolution to generate each signal sample.

In the present, in 7 As shown, a moving eight tap average filter is used such that the output sample is generated from an average of the values within the eight tap window. As it is briefly explained in other embodiments, contains a low-pass filter 36 as a portable eight tap averaging filter a portable four taps averaging filter and two mobile two tap averaging filters. These are used to provide an increased passband of the low pass filter as the details of the reproducing samples move to the central tap of the eight tap filter. This causes a higher level of filtered extended signal samples, where the detector processor determines that these signal samples represent a detail.

8th provides a representation of the low pass filter 36 after a sample execution. In 8th has the low pass filter 36 three moving average filters, which have eight tap filters with one in 7 include the impulse response shown. As shown in 8th is an eight tap delay part 50 used to delay the input signal samples to eight samples. At the output of the eighth tap are adders 52 . 54 used to sample the output in combination with a single-stage register 56 to create. Effectively realized in 8th the filter shown a moving average filter by the accumulation of the last eight signal samples, as by the adder 54 while the last value is subtracted to pass through and outside the eight tap window, as by the adder 52 executed falls. The adders 52 . 54 and the register 56 form an accumulation unit for this purpose 58 to perform the calculation of the output-side mean value.

In the 8th shown accumulation unit 58 thus effectively implements a moving average filter without the need for summing and splitting all of the signal sample components within the eight tap window to calculate each output signal sample. The offers in 8th The circuit shown has the advantage of reducing the number of components needed to realize the moving average filter which does not require multipliers. Correspondingly, the accumulation units form 58 ' . 58 '' the output signal samples associated with the moving four tap and two tap averaging filters. Finally, the second accumulator 58 ' configured to receive samples from the second stage and the sixth stage to form the movable four tap filter, while the third accumulator 58 '' is set up to receive samples from the third and fifth stages. Thus, the eight, four and two tap filters are centered on a common reference.

As in 8th shown, the control signal from the channel 38 through the selection processor 59 receive. The selection processor 59 has four input channels 51 . 53 . 55 . 57 , The four input channels 51 . 53 . 55 . 57 are set up to read the output sample from the moving eight tap averaging filter 51 , the output side sample from the mobile four tap averaging filter 53 , the output side sample from the mobile two tap averaging filter 55 and the signal sample from the fourth stage of the shift register 50 receive accordingly. Selection of the output of either the two tap, four tap or eight tap output, or bypassing the sample from the fourth stage of the shift register is dependent on the received control signal from the detector processor 34 certainly.

To avoid the likelihood of reducing the clarity of such details as corners or individual pixel objects in the rendered image, the detector processor is employed to detect such details. Then, the detector processor generates a control signal to control the output side Select samples from either the mobile eight tap, four tap, or two tap averaging filters, or let the sample pass. The selection signal effectively adjusts the bandwidth of the low-pass filter as a function of the relative position of the pixel associated with the detail relative to the center. 9 provides an example implementation of the detector processor 34 , As shown in 9 includes the detector processor 34 a shift register 60 which has eight stages, the stages of the shift register used in the low-pass filter 50 assigned. As in 9 shown is an adder 62 configured to receive an input signal sample and a signal sample from the second stage output of the shift register 60 to recieve. The adder 62 is arranged to subtract the newly received signal sample from the signal sample present in the second stage by a difference at the output 64 issue. A threshold processor 66 thereafter compares the difference between the signal samples with a predetermined threshold, and if the difference is greater than the threshold, the processor generates 66 a control signal on the control channel 38 , The threshold is determined so that a detail is detected when all three RGB signal samples are greater than the associated RGB values of a pixel from two previous sample periods. Still, in addition to the adder 62 another adder 68 configured to calculate a difference between the output of the fourth stage and the second stage, which is also the threshold processor 66 is supplied.

The operation of the detector processor is illustrated by a flowchart in FIG 10 shown. As shown in 9 the latest signal sample is compared with the contents of the second previous signal sample, as by the adder 62 executed. The threshold processor 66 determines whether the detail between these two compared signal samples has been detected, which is in 10 as the step S2 is shown.

As mentioned above, the threshold processor receives 66 a comparison between the output of the second stage of the third shift register and the output of the fourth stage of the shift register, as by the adder 68 executed. Therefore, the threshold processor determines through step S4 in the flow chart whether the detected detail was really between the latest signal sample and the second previous signal sample within a four tap window provided by sample values at the outputs of the third, fourth, fifth, and sixth register stages.

If the threshold processor determines that the detail is not within a four tap filter has been detected, but in the eight tap Filter exists by comparing the contents of the second and fourth stages of the eight tap delay, then the threshold processor determines in step S4, that the four tap filter should be selected to the output side To generate pixel samples in step S5. If the threshold processor determines that the detail within a four tap filter, but has been captured just before the two central taps, which is the two Tap filter, the process proceeds to step S8.

In step 8, the value of the signal sample in the fourth stage is also compared to a second threshold, which is determined between the threshold for the detection of details and a predetermined offset value, as expressed below:
Threshold <sample value <(threshold + offset value)

If the signal sample value is greater than the threshold but less than the threshold plus the offset value, then the control signal is the output of the moving two-tap average filter 55 to choose on the channel 38 generated as performed in step S10. However, if the value of the signal sample is greater than the threshold plus the offset, then the signal sample will bypass the output of the fourth tap of the shift register 50 selected.

If the detail has not been detected in either step S2 or S4, then the process continues by loop S6 to the beginning of step S2 and it becomes the control signal for selecting the output of the eight Tap filters generated.

One Another advantage is given by the choice of either of the four Tap output or two tap output regardless of the relative position a sample associated with a detail in the color image, and the Value of the samples representing the color image. In fact, the Threshold processor adjusts the bandwidth of the low pass filter dependent on from the relative frequency bandwidth of the pixel samples passing through pass the filter. Accordingly, since the detail pixel samples within the eight tap window, but just before the four tap window then the four-tap output is selected. If the detail pixel samples within the four tap window, but short before the two tap window has been detected, then the output of the two tap windows selected. In other cases will be the output side samples of the mobile eight tap averaging filter selected. By selecting a degree of filtering as a function of the relative Position in the eight tap window will be an improved playback implemented the color image, in which the probability of Contour effects is reduced.

Further it was found that the quality of the color image is better, if the moving two tap averaging filter to generate the output side Samples are used when the details assigned in the color image Samples are determined between the threshold and the threshold plus predefined offset value to fall. The predetermined offset value is set to capture the sample values that exceed the threshold exceed with a relatively small value. It has been found that by filtering out such sample values with a mobile two tap averaging filter an improvement the reproduced color image is effected.

This is because those sample values just exceeding the threshold but not the offset value are associated with the detail centered in the reproduced color image, which requires a somewhat limited degree of filtering with respect to the two immediate samples. Therefore, as illustrated in steps S7, S8 and S10, if the difference between signal samples is greater than the threshold plus the offset value, then the filter is bypassed and the central sample is that in the fourth stage of the shift register 60 contained, passed to the output of the filter.

A result of processing a 12-bit signal sample, in 3 shown image, with the image processor 20 is in 11 to see. As in 12 can be seen, the contour effects that are in the face of the woman in 3 appear, removed. Nevertheless, the fine detailed information in this picture has been preserved.

Even though in the example described above, the low-pass filter for filtering of only one-dimensional signal samples can be set up in alternative versions A two-dimensional filter can be used for further improvement in the displayed image. It is therefore noted that the filter could be set up to vertical and horizontal To filter component samples of the image.

Various Modifications can with the versions of the invention hereinbefore described are without departing from the teachings of the invention. In particular, how can be noted, a color image by color samples with a gar smaller number of bits. The color image can then with increased Bit resolution be restored as the color signal samples from the memory read to the original resolution according to the above described embodiments. Thus, it should be noted that resolutions other than 12 bits are used to store the Color image, and the samples with extended resolution of 18 bits considered can be.

Claims (17)

An image processing apparatus operable to process input color component signal samples representing an input color image to produce output color component signal samples representing an enhanced version of the input color image, the apparatus comprising: an expansion processor ( 20 ) operable to extend the dynamic range of the color component signal samples by increasing the number of bits used to represent each sample, a low-pass filter (US Pat. 36 ) operable to filter the extended dynamic range color signal samples to form the output-side color component samples, and a detector processor (Fig. 34 ) operable to detect color component signal samples associated with a detail representation appearing in the input-side color image, the detector processor being operable to detect the bandwidth of the low-pass filter (Fig. 36 ) is adjustable in accordance with the detected detail. An image processing apparatus according to claim 1, wherein the detector processor ( 34 ) is operable to the bandwidth of the low-pass filter ( 36 ) by selecting the color samples representing the detail with the extended dynamic range, without passing through the low-pass filter ( 36 ). An image processing apparatus according to claim 1 or 2, wherein the input side color signal samples are represented as a predetermined number n of bits, and the expansion processor (16). 20 ) is operable to increase the resolution of the input side color component samples by adding a predetermined number l of additional least significant bits to the input color signal samples and by copying the corresponding number l of most significant bits of the input component signal samples into the 1 least significant bits. Image processing apparatus according to claim 1, 2 or 3, wherein the low-pass filter ( 36 ) includes a moving average filter having a predetermined number of taps d. An image processing apparatus according to claim 4, wherein the input-side color component samples pass through the moving average filter to form the output-side color component samples, following the detection of the detail, the samples of the input-side color component samples associated with that detail Exit when the assigned samples reach a central tap of the moving average filter. Image processing device according to one of the preceding claims, wherein the low-pass filter ( 36 ) comprises a plurality of low-pass filters, each of which is provided with a different bandwidth, and the detector processor is operable to adjust the bandwidth of the low-pass filter by selecting the output signal of one of these low-pass filters. An image processing apparatus according to claim 6, wherein each of the low-pass filters comprises a moving average filter, each of which has a different number of taps, wherein each of the moving average filters has a different window length provides input-side signal samples that are related to a common temporary Sample position is centered, and the detector processor for detecting is operable by samples for considering the detail the common temporary Reference representative and, following acquisition, the output signal sample from an associated moving average filter before selecting the detail samples have entered the window of the filter. Image processing apparatus according to claim 6 or 7, wherein the detector processor ( 34 ) for detecting the samples associated with the detail by comparing successive samples present in the taps of the moving average filters. Image processing apparatus according to claim 8, wherein the detector processor ( 34 ) for detecting the samples associated with the detail by comparing each color component sample of one pixel with the associated color component sample of a subsequent pixel, wherein when each color component sample of the subsequent pixel has changed, the detail represented by the subsequent pixel is detected. Image processing apparatus according to claim 8 or 9, wherein the compared samples separated by two sample periods are. Display device containing a display ( 2 ), a display memory ( 1 ) and an image processing device according to one of claims 1 to 10, wherein input side color component samples representing a color image are stored in the display memory ( 1 ) are stored before their display, and the input color component samples from the display memory ( 1 ) and processed by the image processing device before being displayed on the display device. A display device according to claim 11, comprising a Compression processor that is operable to input side Receive color component samples and reject the resolution of each sample to reduce at least one least significant bit of each sample, wherein the samples having the reduced resolution are stored in the memory be, with the for the storage of the color image required capacity of the memory corresponding to the number of discarded least significant bits each Sample is reduced. A display device according to claim 11 or 12, wherein the display is a liquid crystal display (LCD) (LCD). 2 ). A display device according to claim 11, 12 or 13, wherein the extended color component signal samples generated by the image processing device are converted to analog form by an analog-to-digital converter ( 12 ) for the display on the LCD display. Computer or communication device with a Display device according to one of claims 11 to 14. Mobile radio telephone with a display device according to one of the claims 11 to 14. Method of processing an input color image representing input-side color component signal samples to an extended version the input side color image representing output side To produce color component signal samples, the method comprising: Increase the dynamic range of the input color component signal samples by raising the number of people to represent bits used in each sample, Low pass filtering the color component signal samples with the extended dynamic range to the output color component samples and collecting color component samples which are one in the input image color image appearing detail representation and following the acquisition, adjust the bandwidth the low-pass filter according to the detected detail.