CN111818319B

CN111818319B - Method and system for improving display quality of three-dimensional image

Info

Publication number: CN111818319B
Application number: CN201910285154.2A
Authority: CN
Inventors: 李应樵
Original assignee: Shenzhen Visual Sense Power Technology Co ltd
Current assignee: Shenzhen Visual Sense Power Technology Co ltd
Priority date: 2019-04-10
Filing date: 2019-04-10
Publication date: 2022-05-24
Anticipated expiration: 2039-04-10
Also published as: CN111818319A

Abstract

The invention discloses a method for improving quality when displaying a 3D image by using a source depth map as input and automatically rendering a three-dimensional (3D) display with two or more viewpoints. The invention comprises a method of mapping (adjusting) depth values to depth coefficients of actual physical deviations, mapping to substantial deviation values, and obtaining a target depth map. Accordingly, the invention corrects the disclosed real-time depth map in the three-dimensional display device and the content production process. The invention can adjust the depth of the 3D image for the three-dimensional (3D) display with different parameters in a targeted way, so that a user feels more comfortable when watching the three-dimensional television.

Description

Method and system for improving display quality of three-dimensional image

Technical Field

The present invention relates to the field of image processing technology, and in particular, to a method and system for improving display quality through naked eye three-dimensional (3D) image depth value adjustment and multi-view generation adjustment.

Background

The traditional image only has two-dimensional (2D) information, and depth information of an object is neglected in the display content, so that the three-dimensional effect is lacked, and the depth level cannot be displayed. With the development of science and technology, three-dimensional information stereoscopic display gets more and more people's favor. The 3D technology has become mature and has entered the application stage, wherein the naked eye 3D display technology gets rid of the constraint of the auxiliary visual devices (such as helmet display, stereoscopic polarized glasses, etc.), and can directly view the stereoscopic effect, so that the user can really experience the stereoscopic vision scene in the scene.

Display depth is an important characteristic of naked-eye stereoscopic displays, which represents an acceptable range of depth values. The multi-view content required by the stereoscopic display can be obtained by a view map and a depth map or disparity map corresponding to the view map through a depth map or disparity map-based rendering (DIBR) method, and the depth map determines the display depth of the corresponding detail in the stereoscopic display. When using depth information to perform 3D rendering, the rendering process performs multi-view generation according to depth data, and in summary, a depth value larger than a zero depth plane (zero plane) represents that the range image is in front of the 3D screen, a depth value smaller than the zero depth plane represents that the range image is in back of the 3D screen, and a depth value equal to the zero depth plane represents that the range image is on the 3D screen. In order to generate images equal to the front and rear positions of the screen, various three-dimensional display devices project different multi-view images to two eyes of a viewer in different methods, so that different images of left and right eyes can be obtained when the viewer views the images. The method is generally applicable to providing a technique for multi-view or two-view based three-dimensional rendering with depth data.

Generally, the existing 3D display provides the user with the option of adjusting the depth, which is called "depth adjustment magnification", and each display has its own defined adjustable depth range, for example, the 3D screen defines the depth adjustment magnification as 0-200 and the preset value as 100. 100 represents × 1.0 magnification, 0 represents × 0.0 magnification (i.e., 2D), and 200 represents × 2.0 magnification. When the user adjusts the parameters, the overall depth data will increase or decrease linearly, as shown in FIG. 1, 130%, 100% and 70%. However, this method reduces the 3D depth as a whole, and cannot adapt to each video, and the original area where the depth does not need to be reduced is also reduced.

Another parameter is depth compensation, i.e. the zero depth plane position, usually at the screen position. The picture offset is zero when the input depth value coincides with the zero depth plane. This means that the images seen by the left and right eyes of the viewer will be the same. The user can change this parameter to bring the zero plane out or into the screen to provide more or less pop-up and 3D effects. In addition, because the display depth of field of the 3D display is affected by the parameters of the 3D display (especially, the parameters such as the grating structure and the focal length of the lenticular lens), the assembling process, the processing error, and the like, the display effect is different when the same video content is played.

Therefore, a new technology is required to solve the above-mentioned problems in the prior art.

The invention mainly aims at the condition that the part with relatively strong depth of the watching part feels blurred due to the visual error caused by the overlarge parallax of different viewpoint images watched by two eyes.

The invention mainly aims to find the maximum acceptable strength of each three-dimensional display device and carry out depth limitation without losing precision by utilizing the adjustment of the depth map, thereby improving the display quality.

Disclosure of Invention

In order to improve the definition of a 3D image, the present invention provides a method and an apparatus for improving display quality by adjusting a depth value of a 3D image. The method for improving the display quality of a three-dimensional (3D) image comprises the following steps: obtaining and inputting a source depth map of the 3D image; setting screen parameters of a 3D display, and obtaining and setting upper and lower limits of depth values acceptable by the screen; adjusting the source depth map according to the screen parameters and the upper limit and the lower limit of the depth value acceptable by the screen; mapping the source depth map to a target depth map according to the adjusting step. The screen parameters are a zero plane depth value of the 3D display and a depth adjustment multiplying factor of the 3D display. The step of obtaining and setting the upper limit and the lower limit of the depth value acceptable by the screen further comprises the following steps: obtaining the upper limit and the lower limit test results of the depth values; calculating an upper depth limit value and a lower depth limit value according to the upper depth value limit test result and the lower depth value limit test result; obtaining a depth coefficient according to the obtained upper limit value and lower limit value of the depth; acquiring upper and lower limits of an actual offset distance according to the depth coefficient; wherein: the substantial offset distance is depth coefficient x (input depth value-zero depth plane depth value); calculating the depth upper limit depth value as the total number of test frame numbers/test frame numbers, the positive depth range value and the zero depth plane depth value; the depth lower depth value is calculated as the negative depth range value-test frame number/total test frame number x zero depth plane depth value. Wherein the depth coefficient is an upper depth limit deviation percentage and a lower depth limit deviation percentage; wherein the upper-limit depth deviation percentage is (upper-limit depth value-zero depth plane depth value)/(maximum depth value-zero depth plane depth value) × original substantially upper-limit deviation percentage; the lower-limit depth deviation percentage is (zero depth plane depth value-lower-limit depth value)/(zero depth plane depth value) multiplied by the original substantial lower-limit deviation percentage. Wherein the source depth map is adjusted using at least one of a clipping correction, a two-segment linear mapping, or a sigmoid (sigmoid) function correction step. Setting a maximum and a minimum depth threshold, wherein the part exceeding the maximum threshold is directly set as the maximum threshold, and the depth value smaller than the minimum threshold is directly set as the minimum depth threshold; the depth values are limited to a selection range. Wherein the depth value to offset conversion equation is：F(Z_i)＝(Z_i-Z_o)*D_f*D_t(4) At F (Z)_i)>In the case of H, H is assigned to F (Z)_i) I.e. F (Z)_i) H; at F (Z)_i)<In the case of H, if F (Z)_i)<L, assigning L to F (Z)_i) I.e. F (Z)_i)＝L；Z_iTo input depth values, D_fThe offset coefficient is the coefficient of offset of each point depth value on the screen; d_tIs the depth magnification. Wherein the two-segment linear mapping comprises the following steps: for the medium and low range values, linear behavior is realized, and for the high range values, coordinates corresponding to 5 points are required to be obtained, wherein the coordinates comprise maximum negative parallax deviation, maximum positive parallax deviation, zero depth plane positions, negative parallax curved nodes and positive parallax curved nodes; wherein

Negative disparity curve node (x, y) ═ Z_g*p,L*q) (5)

Positive disparity warp node (x, y) ═ 255- (Z)_g*p),H*q) (6)

Depth value (Z) of zero depth plane_g) Preset to 128. And p is an X-axis curve node proportion coefficient and is used for calculating an X coordinate of an input curve node. And q is a Y-axis curve node proportion coefficient and is used for calculating the Y coordinate of the output curve node. Wherein the S-type (sigmoid) function modification step comprises the following steps: determining a sigmoid function through a sigmoid equation:

wherein a, b and c are variables to be determined, D_outIs the target depth map output value, D_inThe method is characterized in that the method is a source depth map input value, e is a natural constant, and theta is the slope of a sigmoid curve. Where θ is 0.03.

The system for changing the display quality of a 3D image of the present invention comprises: the input module is used for obtaining and inputting a source depth map of the 3D image; the setting module is used for setting screen parameters of the 3D display, and obtaining and setting an upper limit and a lower limit of depth values acceptable by a screen; the adjusting module is used for adjusting the source depth map according to the screen parameters and the upper limit and the lower limit of the depth value acceptable by the screen; and the mapping module maps the source depth map into a target depth map according to the adjusting step.

In yet another aspect of the present invention, one or more computer-readable media storing computer-executable instructions that, when used by one or more computer devices, cause the one or more computer devices to perform the method of changing the display quality of a 3D image are also provided.

The video 3D quality adjusted according to the invention is better, more details are retained, and the contrast is improved, thereby improving the experience effect of the user. The invention can carry out targeted processing on the 3D video data, so that different displays can have better 3D effect when the same video content is played.

Drawings

Fig. 1 is a diagram illustrating the corresponding deviation relationship between a depth map input and multiple views in an unmodified condition.

Fig. 2 shows a flow chart of a method of adjusting depth values of a 3D image according to the invention.

Fig. 3 shows a schematic diagram comparing depth map maps with different upper and lower limits according to an embodiment of the invention.

Fig. 4 shows a comparative schematic of a depth map with different upper and lower limits according to another embodiment of the invention.

Fig. 5 shows a comparative schematic of a depth map with different upper and lower limits according to yet another embodiment of the present invention.

Fig. 6 illustrates a schematic diagram of a 0 th frame of a test depth parallax video according to an embodiment of the present invention.

Fig. 7 shows a schematic diagram of a 185 th frame of a test depth disparity video according to an embodiment of the present invention.

Fig. 8 illustrates a schematic diagram of a 370 th frame of a test depth disparity video according to an embodiment of the present invention.

Fig. 9 illustrates a block diagram of a system for adjusting depth values of a 3D image according to an embodiment of the present invention.

FIG. 10 schematically shows a block diagram of a server for performing the method according to the invention; and

fig. 11 schematically shows a storage unit for holding or carrying program code implementing the method according to the invention.

Detailed Description

In order to make the technical solutions of the present invention better understood, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are only a part of examples of the present invention, and not all embodiments. All other embodiments, which can be obtained by a person skilled in the art without any inventive step based on the embodiments of the present invention, shall fall within the scope of the present invention.

Fig. 1 is a schematic diagram of the corresponding deviation relationship of the depth map input and the multiple views in the unmodified condition. When the user adjusts the parameters, the overall depth data will increase or decrease linearly. As shown in fig. 1, in the unmodified condition, the corresponding deviation relationship of the depth map input and the multi-view is linear. The user can change the depth magnification, and the figure shows 3 parameter values, 130%, 100% and 70%, respectively.

Fig. 2 shows a flow diagram of a method 200 of adjusting depth values of a 3D image according to an embodiment of the invention.

In step 202, a source depth map of the 3D image is obtained and input. A depth map, i.e. a source depth map, typically input to a 3D display, the value of which is 0-255 if the depth value is 8 bits per pixel, wherein the number of bits is 128; if the depth value is 16 bits per pixel, the value of the depth map is 0-65535, where the number of bits is 32768.

In step 204, screen parameters of the 3D display are set and upper and lower limits of depth values acceptable to the screen are obtained and set. The screen parameters include two parts: 1.zero plane depth value of 3D display; and 2.3D display depth adjustment multiplying power. Wherein the zero-plane depth value refers to the position of the zero-depth plane on the screen, and in the first example of the source depth map, the zero-depth plane value is 128; in the example of fig. 1, the depth adjustment magnification of the 3D display is a magnification of × 1.0, i.e., a depth magnification of 100%. Due to limitations of the display device and lenticular lens (e.g. their manufacturing quality), some part of the range of depth maps cannot be fully applied to displaying 3D images. Good 3D display effect has the requirement of strict matching with the depth information value of the display screen. It is therefore necessary to determine the above-mentioned parameters of the 3D display used and to adjust the depth map accordingly on the basis of these parameters. These parameters may be provided by the 3D display manufacturer and are preferably measured for each display after the manufacturing process.

The maximum depth shift is defined as the maximum shift in microns for the left and right eye images on a single image. This is a measurable parameter and has a direct relationship to the viewer's viewing comfort. Different designs have different acceptable maximum depth offsets for screen viewing distances. The actual offset distance is calculated as:

virtual offset distance ═ depth coefficient × (input depth value-zero depth plane depth value) (1)

A method of obtaining acceptable maximum and minimum depth values (i.e., disparity indexes) for a 3D display according to an embodiment of the present invention will be described below in conjunction with fig. 6-8.

Fig. 6 is a diagram of frame 0 of a depth parallax video tested according to an embodiment of the invention. The depth of all dots is zero depth plane depth value.

FIG. 7 is a 185 th frame of a depth-disparity video tested according to an embodiment of the present invention. The depth of all dots is half of the dynamic range.

Fig. 8 is a diagram of 370 th frame of a depth parallax video tested according to an embodiment of the present invention. The depth of all dots is the maximum of the dynamic range.

In the example of fig. 6-8, the leftmost first and third rows are upper measurement depth limits and the leftmost second and fourth rows are lower measurement depth limits. The upper limit of measurement is the first and third rows, and the tester needs to perform four tests. The first test observer needs to watch the first row, watch the test movie and pause when watching this perceived as a blurry, unclear, uncomfortable spot detection of the first row, pause the test film source and record the numbers below. The second to fourth tests are carried out by watching the second to fourth rows of dots in the same way and respectively recording the numbers during pause.

The measurement image has the lowest depth value at frame 0 and the highest depth value at frame 370, and if the depth value uses 8 bits (bit), the values of the depth map are 0-255 values, and the median is 128, i.e. a zero depth plane. The positive depth range is 129-255, with a positive depth range value of 127. The negative depth range is 0-127 with a negative depth range value of 128. And estimating the upper and lower depth limits of comfortable watching according to the parameter frame number recorded by the tester. The upper and lower limits are defined as:

upper depth limit is the average of the first and third row test results/370 x positive depth range value + zero depth plane depth value (2)

Lower depth limit-negative depth range value (mean value of test results in rows 2 and 4/370 × zero depth plane depth value) (3)

According to the obtained upper and lower limits of the depth, the substantial deviation upper and lower limits can be obtained through the formula (1), and the adjustment of the upper and lower limits of the practical depth can be achieved when the depth multiplying power is adjusted.

The following are examples of calculations to obtain acceptable maximum and minimum depth values in the examples shown in FIGS. 6-8:

as described above, fig. 6, 7 and 8 illustrate video frames of a method for measuring maximum and minimum depth values acceptable for a 3D display according to an embodiment of the present invention, in which fig. 6 is a 0 th frame of a video, fig. 7 is a 185 th frame of a video, and fig. 8 is a 370 th frame of a video. In fig. 8, the first column and the third column denoted by reference numeral 701 are used to test positive parallax, with depth values increased from 128 to 255. The second and fourth columns, denoted by reference numeral 702, are used to test for negative disparity, with depth values reduced from 128 to 0. First, the depth coefficient is adjusted to a maximum value and the depth offset is adjusted to 128 until the grid lines on the video are completely clear. The tester then stands at the optimal viewing position for which the display has been designed, plays the video and observes the points on the left and right screens. The time and location of the point at which the tester cannot perform stereo vision is recorded, i.e. 2 points are seen instead of 1 point. And finally, calculating the acceptable maximum and minimum depth parameters according to the time and the position of the point, finding the parallax index and recording the parallax index.

In step 206, the source depth map is adjusted according to the screen parameters and the upper and lower limits of acceptable depth values of the screen, and in step 208, the adjusted source depth map is mapped to the target depth map. The present invention proposes the following adjustment methods, and steps 206 and 208 may map the source depth map using one or more of the adjustment methods.

Fig. 3-5 are schematic diagrams illustrating comparison of depth map maps with different upper and lower limits according to different embodiments of the present invention. FIG. 3 is a schematic diagram of the corresponding deviation relationship between the input depth map and the multi-views using the cropping correction method in the range of the upper limit of the 6% offset and the lower limit of the-5% offset. FIG. 4 is a diagram illustrating the corresponding deviation relationship between the input depth map and the multiple views using the two-step linear mapping method in the range of the upper deviation limit of 6% and the lower deviation limit of-5%. FIG. 5 is a diagram illustrating the corresponding deviation relationship between the depth map input and the multi-views using the sigmoid function correction method in the range of the upper limit of the 6% offset and the lower limit of the-5% offset. The invention provides three adjusting methods: hard clipping, two-stage linear mapping and sigmoid function mapping. sigmoid function mapping is relatively complex compared to the other two. Users can weigh this, depending on their application. As shown in fig. 3-5, compared with the hard clipping and two-segment linear mapping method, the sigmoid function mapping method is smoother, can retain details related to the source depth to the maximum extent, and has the best display effect.

As shown in FIG. 3, the first embodiment of the adjustment method according to the present invention may also be referred to as hard clipping, where the depth values are limited to selectionA range. A positive (H) and negative (L) upper excursion limit is first calculated from the acceptable maximum and minimum depth values. Input depth value (Z)_i) To mapping equation F (Z)_i) The output is the deviation value of the Y-axis in millimeters (mm).

The depth-to-offset conversion equation is: f (Z)_i)＝(Z_i-Z₀)*D_f*D_t (4)

At F (Z)_i)>In the case of H, H is assigned to F (Z)_i) I.e. F (Z)_i) H; at F (Z)_i)<In the case of H, if F (Z)_i)<L, assigning L to F (Z)_i) I.e. F (Z)_i)＝L；Z_iTo input depth values, the depth values use 8 bits (bit), and the values of the depth map are 0-255 values. Zero plane depth value (Z)_o) Is 128. D_fIs a shift coefficient, i.e., a coefficient of shift in the screen of the change in depth value per point. D_tFor depth magnification, the user can adjust the depth magnification to be 70%, 100% and 130%, respectively. As shown in fig. 3, the maximum and minimum depth thresholds are set, and the portion exceeding the maximum threshold is directly set as the maximum threshold, and the depth value smaller than the minimum threshold is directly set as the minimum depth threshold.

As shown in fig. 4, another embodiment of the adjustment method according to the present invention can also be referred to as a two-segment linear mapping method. The two-segment linear mapping is characterized by linear behavior for mid-low range values, while for high range values it provides a smoother solution than the shearing method described in fig. 3. During calculation, the coordinates corresponding to the 5 points are required to be obtained, and the coordinates comprise maximum negative parallax deviation, maximum positive parallax deviation, zero-depth plane position, negative parallax curved node and positive parallax curved node. The result of FIG. 4 is obtained by connecting 5 points.

Negative disparity curve node (x, y) ═ Z_g*p,L*q) (5)

Positive disparity warp node (x, y) ═ 255- (Z)_g*p),H*q)(x,y)＝(255-(Z_g*p),H*q) (6)

Wherein the depth value (Z) of the zero depth plane_g) Preset to 128. And p is an X-axis curve node proportion coefficient and is used for calculating an X coordinate of an input curve node. q is Y-axis curved node proportion systemAnd the number is used for calculating the y coordinate of the output curve.

In one specific example, the q-scaling factor is p-0.4 q-0.9. Assuming that the mapping applies to the input depth map and image for the most part at depth values (128x0.4 for 50 and 256 for 128x0.4 for 205), the regions at the range at the hierarchical level near the upper and lower limits are compressed, thus making the image at intermediate depth values more distinct. One of ordinary skill in the art may make functional modifications as needed to particularly emphasize the desired depth portion. The above example shows the function F as a preset when the depth coefficient adjustment magnification is x1, and fig. 4 demonstrates the effect when the magnification is 0.7 and 1.3.

As shown in fig. 5, a sigmoid (S-shaped) mapping function method is proposed in the further embodiment of the adjusting method according to the present invention to adjust the depth map of the 3D image, and as shown in fig. 5, compared with the hard cropping and two-segment linear mapping method, the depth map mapping is smoother, the details related to the source depth map can be retained to the maximum extent, and the display effect is the best.

In step 206, a sigmoid function is determined based on the obtained zero-plane depth values, acceptable maximum and minimum depth values. To determine the sigmoid function, the sigmoid equation is as follows:

wherein a, b and c are variables to be determined, D_outIs the target depth map output value, D_inThe method is characterized in that the method is a source depth map input value, e is a natural constant, and theta is the slope of a sigmoid curve.

To solve the above equation, for the source depth value input range [0, 255], three transformation equations may be used, respectively related to the zero plane depth value, the acceptable maximum and minimum depth values of the 3D display:

wherein, T_upperRepresenting an acceptable maximum depth value, T_lowerRepresents the minimum acceptable depth value, (Z)_x,Z_y) Representing the zero plane of the depth map.

To solve the above system of equations, the value of θ is also determined. If the other variables in equation (7) are unchanged, the slope of the sigmoid mapping function increases as θ increases. From the experimental results, as shown in fig. 5, when θ is 0.03, the smoothness of the Sigmoid curve is optimal, and thus it can be selected to be the value of θ required to optimize the depth map; of course, different values of θ may be selected according to the curve requirements, such as slope. Substituting θ into the above equation 3-5 with 0.03 yields:

the sigmoid function depth maps the modified equation with zero plane depth value, acceptable maximum and minimum depth values for the 3D display set.

First, a positive (H) and a negative (L) upper excursion limit are calculated from the acceptable maximum and minimum depth values, the depth value (Z) of the zero depth plane_g) Preset to 128. Input depth value (Z)_i) To mapping equation F (Z)_i) The output is the deviation value of the Y axisIn millimeters (mm).

Mapping equation

Theta controls the shape of the curve. The larger the value of θ, the higher the slope, preset to 0.03. And is also defined as a depth coefficient variable value, i.e. representing θ × depth coefficient. The curve shape is changed under different depth coefficients. Regarding the consideration of setting θ, the reason why the present example sets θ to a value of 0.03 is that when the zero depth plane of the input depth is 0, the slope of the mapping equation curve is 1. This slope is the same as the overall slope of the depth and offset relationship equation without any modification. The desired theta function may be selected in consideration of the desired shape of the mapping curve when designing using the above equation.

A typical three-dimensional display device provides adjustable depth coefficients, which typically are tilt coefficients that modify a depth mapping curve, which may be modified by a sigmoid mapping method to change the tilt of the curve to a near zero depth plane. Fig. 5 shows the results of θ values x1.3 and x0.7 times. As shown in fig. 5, it has a smooth transition function.

In step 208, the source depth map of the 3D image is mapped to the target depth map using the determined sigmoid function, thereby obtaining a better image that can be displayed on a given 3D display.

In the present invention, the S-shape of the sigmoid mapping function assists the dynamic range mapping process in two ways: 1) approximately preserving the original perceived depth over a wide range of destination dynamic ranges by increasing image contrast during the remapping process; 2) the low end compression is gradual, thereby reducing the hard cut low end depth deficiency. The invention determines the maximum and minimum depth values that can be accepted by the parameters of the 3D display used, such as the zero plane, and adjusts the depth map of the 3D image according to the obtained parameters. The adjusted video has better 3D quality, more details are reserved, and the contrast is improved, so that the experience effect of a user is improved. The method can process the 3D video data in a targeted manner, so that different displays can have better 3D effect when the same video content is played.

As shown in fig. 9, the present invention also provides a system 900 for adjusting depth values of a 3D image, the system 900 comprising: an input module 902 configured to obtain and input a source depth map; a setting module 904 configured to set screen parameters of the 3D display and obtain and set upper and lower limits of depth values acceptable for the screen. The screen parameters include two parts: 1.zero plane depth value of 3D display; 2, adjusting the depth of the 3D display by multiplying power; an adjustment module 906 configured to adjust the source depth map according to the screen parameters and the acceptable maximum and minimum depth values; can be prepared by at least one of the following methods: adjusting a method for determining a sigmoid function by a hard cutting method, a two-segment linear mapping method and an S-shaped mapping function; and a mapping module 908 configured to map a source depth map of the 3D image to a target depth map using functions determined by the at least one adaptation method, including a clipping function, a two-segment linear function, and a sigmoid function. Wherein the sigmoid mapping function in the adjustment module 906 is further configured to provide a proposed sigmoid mapping equation:

wherein the positive (H) and negative (L) deflection limits, the depth value (Z) of the zero depth plane_g) Preset to 128, e is a natural constant, and theta is the slope of the sigmoid function. Input source depth map input value (Z)_i) To mapping equation F (Z)_i) And outputting the deviation value, wherein the deviation value is in millimeter (mm) as a unit, the zero plane depth value, the acceptable maximum depth value and the acceptable minimum depth value of the 3D display are respectively substituted into the sigmoid equation to obtain a transformation equation set so as to determine a sigmoid function, and 3D rendering is performed according to the output deviation value, so that a user feels more comfortable when watching the three-dimensional television.

Furthermore, the present invention also provides one or more computer-readable media storing computer-executable instructions that, when used by one or more computer devices, cause the one or more computer devices to perform a method of adjusting depth values of a 3D image according to the present invention. The computer readable media may be any available media that can be accessed by the computer device and includes both volatile and nonvolatile media, removable and non-removable media implemented in any method or technology for storage of information such as computer readable instructions, data structures, program modules or other data. Computer-readable media includes, but is not limited to, RAM, ROM, EEPROM, flash memory or other memory technology, CD-ROM, Digital Versatile Disks (DVD) or other optical storage, magnetic cassettes, magnetic tape, magnetic disk storage or other magnetic storage devices, or any other medium which can be used to store the desired information and which can be accessed by a computing device. Combinations of any of the above should also be included within the scope of computer readable media. For example, FIG. 10 illustrates a server, such as an application server, in which embodiments in accordance with the present invention may be implemented. The server conventionally includes a processor 1010 and a computer program product or computer-readable medium in the form of a memory 1020. The memory 1020 may be an electronic memory such as a flash memory, an EEPROM (electrically erasable programmable read only memory), an EPROM, a hard disk, or a ROM. The memory 1020 has a storage space 1030 for program code 1031 for performing any of the method steps of the above-described method. For example, the storage space 1030 for program code may include respective program code 1031 for implementing various steps in the above method, respectively. The program code can be read from or written to one or more computer program products. These computer program products comprise a program code carrier such as a hard disk, a Compact Disc (CD), a memory card or a floppy disk. Such a computer program product is typically a portable or fixed storage unit as described with reference to fig. 6. The storage unit may have a storage section, a storage space, and the like arranged similarly to the memory 1020 in the server of fig. 11. The program code may be compressed, for example, in a suitable form. Typically, the storage unit comprises computer readable code 1031', i.e. code that can be read by a processor, such as 1010 for example, which when executed by a server causes the server to perform the steps of the method described above.

The present invention has been illustrated by the above examples, but it should be understood that the above embodiments are for illustrative and descriptive purposes only and are not intended to limit the invention to the scope of the described embodiments. Furthermore, it will be understood by those skilled in the art that the present invention is not limited to the embodiments described above, and that many variations and modifications may be made in accordance with the teachings of the present invention, which variations and modifications fall within the scope of the present invention as claimed. The scope of the invention is defined by the appended claims and equivalents thereof.

Claims

1. A method of improving three-dimensional (3D) image display quality, comprising:

obtaining and inputting a source depth map of the 3D image;

setting screen parameters of a 3D display, and obtaining and setting upper and lower limits of depth values acceptable by the screen;

obtaining the upper limit and the lower limit test results of the depth values;

calculating an upper depth limit value and a lower depth limit value according to the upper depth value limit test result and the lower depth value limit test result;

obtaining a depth coefficient according to the obtained upper limit value and lower limit value of the depth;

acquiring upper and lower limits of an actual offset distance according to the depth coefficient; wherein:

the substantial offset distance is depth coefficient x (input depth value-zero depth plane depth value);

calculating the depth upper limit depth value as the total number of test frame numbers/test frame numbers, the positive depth range value and the zero depth plane depth value;

the depth lower limit depth value is calculated as a negative depth range value-the total number of test frame numbers/test frame numbers multiplied by a zero depth plane depth value;

wherein the depth coefficient is an upper depth limit deviation percentage and a lower depth limit deviation percentage; wherein the upper-limit depth deviation percentage is (upper-limit depth value-zero depth plane depth value)/(maximum depth value-zero depth plane depth value) × original substantially upper-limit deviation percentage;

the lower-limit depth deviation percentage is (zero depth plane depth value-lower-limit depth value)/(zero depth plane depth value) multiplied by the original actual lower-limit deviation percentage;

adjusting the depth coefficient to a maximum value, adjusting the depth offset to 128 until grid lines on the video are completely clear, playing the video and observing points on the left and right screens, recording the time and the position of the point on which the tester cannot perform stereoscopic vision on the point, and calculating the upper limit and the lower limit of the depth value acceptable by the screens according to the time and the position of the point; obtaining a parallax index;

adjusting the source depth map according to the screen parameters and the upper limit and the lower limit of the depth value acceptable by the screen;

mapping the source depth map to a target depth map according to the adjusting step.

2. The method of claim 1, wherein the screen parameters are a zero-plane depth value of the 3D display and a depth adjustment magnification of the 3D display.

3. The method of claim 1, wherein the source depth map is adjusted using at least one of a crop correction, a two-segment linear mapping, or a sigmoid (sigmoid) function correction step.

4. A method as claimed in claim 3, wherein the step of cropping corrections is setting a maximum and a minimum depth threshold, the parts exceeding the maximum threshold being directly set to the maximum threshold, the depth values below the minimum threshold being directly set to the minimum depth threshold; the depth values are limited to a selection range.

5. The method of claim 4, wherein

The depth-to-offset conversion equation is: f (Z)_i)＝(Z_i-Z_o)*D_f*D_t (4)

At F (Z)_i)>In the case of H, H is assigned to F (Z)_i) I.e. F (Z)_i) H; at F (Z)_i)<In the case of H, if F (Z)_i)<L, assigning L to F (Z)_i) I.e. F (Z)_i)＝L；Z_oZero plane depth value; h is the upper limit of positive offset; l is the negative upper limit of excursion; z_iTo input depth values, D_fThe offset coefficient is the coefficient of offset of each point depth value on the screen; d_tIs the depth magnification.

6. The method of claim 3, wherein the two-segment linear mapping step is: for the medium and low range values, linear behavior is realized, and for the high range values, coordinates corresponding to 5 points are required to be obtained, wherein the coordinates comprise maximum negative parallax deviation, maximum positive parallax deviation, zero depth plane positions, negative parallax curved nodes and positive parallax curved nodes; wherein

Negative disparity curve node (x, y) ═ Z_g*p,L*q) (5)

Positive disparity warp node (x, y) ═ 255- (Z)_g*p),H*q) (6)

Wherein the depth value Z of the zero depth plane_gPresetting as 128, p is an X-axis curve node proportionality coefficient used for calculating an X coordinate of an input curve node, and q is a Y-axis curve node proportionality coefficient used for calculating a Y coordinate of an output curve node.

7. The method of claim 3, wherein the sigmoid function modifying step is: determining a sigmoid function through a sigmoid equation:

8. The method of claim 7, wherein θ is 0.03.

9. A system for changing the display quality of a 3D image, the system comprising:

the input module is used for obtaining and inputting a source depth map of the 3D image;

the setting module is used for setting screen parameters of the 3D display, and obtaining and setting an upper limit and a lower limit of depth values acceptable by a screen;

the setting module also comprises a setting module and a control module,

the calculation module is used for obtaining the upper limit and the lower limit test results of the depth values;

the depth lower limit depth value is calculated as a negative depth range value-test frame number/total test frame number multiplied by zero depth plane depth value;

a parallax index obtaining module, which adjusts the depth coefficient to a maximum value, adjusts the depth offset to 128 until grid lines on the video are completely clear, plays the video and observes points on the left and right screens, records the time and the position of the point on which the tester can not perform stereo vision, and calculates the upper limit and the lower limit of the depth value acceptable by the screens according to the time and the position of the point; obtaining a parallax index;

the adjusting module is used for adjusting the source depth map according to the screen parameters and the upper limit and the lower limit of the depth value acceptable by the screen;

and the mapping module is used for mapping the source depth map into a target depth map according to the adjusting step.

10. The system of claim 9, wherein the screen parameters are a zero plane depth value of the 3D display and a depth adjustment magnification of the 3D display.

11. The system of claim 9, wherein the adjustment module adjusts the source depth map using at least one of a crop correction, a two-segment linear mapping, or a sigmoid (sigmoid) function correction module.

12. The system of claim 11, wherein the cropping correction module sets maximum and minimum depth thresholds, the portions exceeding the maximum threshold being set directly to the maximum threshold, the depth values below the minimum threshold being set directly to the minimum depth threshold; the depth values are limited to a selection range.

13. The system of claim 12, wherein

The depth-to-offset conversion equation is: f (Z)_i)＝(Z_i-Z_o)*D_f*D_t (4)

14. The system of claim 11, wherein the two-segment linear mapping module is: for the medium and low range values, linear behavior is realized, and for the high range values, coordinates corresponding to 5 points are required to be obtained, wherein the coordinates comprise maximum negative parallax deviation, maximum positive parallax deviation, zero depth plane positions, negative parallax curved nodes and positive parallax curved nodes; wherein

Negative disparity curve node (x, y) ═ Z_g*p,L*q) (5)

Positive disparity warp node (x, y) ═ 255- (Z)_g*p),H*q) (6)

Wherein the depth value Z of the zero depth plane_gThe preset value is 128, p is an X-axis curve node proportion coefficient and is used for calculating an X coordinate of an input curve node, and q is a Y-axis curve node proportion coefficient and is used for calculating a Y coordinate of an output curve node.

15. The system of claim 11, wherein the sigmoid (sigmoid) function modification module is: determining a sigmoid function through a sigmoid equation:

16. The system of claim 15, wherein θ is 0.03.

17. One or more computer-readable media storing computer-executable instructions that, when used by one or more computer devices, cause the one or more computer devices to perform the method of improving the display quality of three-dimensional (3D) images of any one of claims 1 to 8.