WO2013185695A1

WO2013185695A1 - Method and device for processing mobile terminal images

Info

Publication number: WO2013185695A1
Application number: PCT/CN2013/080113
Authority: WO
Inventors: 张静
Original assignee: 中兴通讯股份有限公司
Priority date: 2012-08-30
Filing date: 2013-07-25
Publication date: 2013-12-19
Also published as: CN102915521A

Abstract

A method and a device for processing mobile terminal images are used for increasing the processing speed. The method comprises: performing area-of-interest detection on an image to be processed according to area-of-interest priori information which is obtained in advance; and improving the resolution of a detected area of interest. The device comprises an area-of-interest detection module and a resolution improving module. Compared with the prior art, an area-of-interest detection part fully takes the priori information of a user interested object which is obtained in advance into consideration, and the detection accuracy of the area of interest is improved; and the resolution improving part of the area of interest only improves the resolution of the detected area of interest, so that the processing speed can be increased.

Description

Mobile terminal image processing method and device

Technical field

The present invention relates to the field of image processing, and in particular, to a mobile terminal image processing method and apparatus.

Background technique

In the field of image processing, image enhancement is an important direction. Image enhancement technology mainly studies the original unclear image or enhances certain features of interest, and suppresses non-attention features to improve image quality, enrich information, and enhance image interpretation and recognition.

The existing image resolution enhancement techniques can be roughly divided into two categories. One is based on multiple low-resolution images, and the image is corrected, and then a high-resolution image is obtained by interpolation, reconstruction or learning. The effect of this kind of method depends largely on the motion parameters before the two images, and the computational complexity is generally higher. The second method is to interpolate a single low-resolution image. High resolution image. This method has a wider range of applications because it involves only low-resolution images.

The widely used interpolation algorithms include the nearest neighbor interpolation algorithm, bilinear interpolation algorithm and bicubic interpolation algorithm. The nearest neighbor interpolation algorithm assumes a linear relationship between the pixels to be interpolated and its horizontal neighbor pixels, which often causes significant distortion. The bilinear interpolation algorithm increases the vertical neighborhood pixels to interpolate on the basis of the nearest neighbor interpolation. Pixel point constraints, the interpolation performance is better than the former; compared to the above interpolation algorithm only considers the influence of pixel 4 neighborhood pixels on the interpolation point, the double cubic interpolation algorithm expands the range to 16 neighborhoods, and uses a smoother curve Calculating the color value of the pixel to be interpolated, the interpolation effect is optimal, but the computational complexity is also the highest.

Summary of the invention

The technical problem to be solved by the embodiments of the present invention is to provide a mobile terminal image processing method and apparatus, which improve processing speed.

In order to solve the above technical problems, the following technical solutions are used:

A mobile terminal image processing method includes: The region of interest is detected according to the a priori information of the region of interest obtained in advance; the resolution of the detected region of interest is improved.

Optionally, the method further includes: before the step of performing the region of interest detection on the image to be processed according to the previously obtained region of interest prior information, the prior information of the region of interest is obtained in advance by:

Providing one or more images to the user, each image containing one or more targets, prompting the user to calibrate the target of interest in the provided image;

Calculating the feature parameters of the user-targeted target of interest; storing the feature parameters of the target of the user's interest as a priori information to the current user's standard interest object library.

Optionally, the step of performing the region of interest detection on the image to be processed according to the previously obtained region of interest prior information includes:

Dividing the image to be processed into one or more object regions according to similar features in the image; calculating feature parameters of the object region;

Comparing the feature parameters of the object region with the feature parameters in the target region of the standard region of interest, and determining the object region with the similarity of the standard feature parameter to be greater than the preset threshold as the user region of interest.

Optionally, the step of performing resolution improvement on the detected region of interest comprises: mapping the detected region of interest into an amplification space according to a preset image magnification N, wherein the amplification space is of interest N times the area;

In the magnifying space, according to the original pixel point, the color value of the pixel other than the original pixel point in the magnifying space is calculated by interpolation, and the local resolution of the image is improved.

Optionally, before the step of mapping the detected region of interest into the amplifying space according to the preset image magnification N, the method further includes:

The region of interest is regularized.

Optionally, the step of calculating, according to the original pixel point in the enlargement space, the color value of the pixel other than the original pixel in the enlarged space by using the interpolation method includes: It is assumed that other pixels except the original pixel point in the enlarged space are pixels to be interpolated, and the following processing is performed on each pixel to be interpolated:

Calculating a degree of correlation between the pixel to be interpolated and the M original pixel points around the pixel, and generating an influence factor of the M original pixel points to the pixel to be interpolated;

And calculating a color value of the pixel to be interpolated according to an original color value of the M original pixel points around the pixel to be interpolated and an influence factor of the M original pixel points to the pixel to be interpolated.

Optionally, the step of calculating a correlation degree between the pixel to be interpolated and the M original pixel points in the surrounding, and generating an influence factor of the M original pixel points to the pixel to be interpolated includes:

计算 Calculate the correlation between each original pixel and the pixel to be interpolated by using the following formula:

Where r _∞ represents the degree of correlation between the mth original pixel point and the pixel to be interpolated, where K m < M , i, j are the horizontal and vertical coordinates of the mth original pixel point, respectively, x and y are The horizontal and vertical coordinates of the pixel to be interpolated;

Performing the following normalization processing on the degree of correlation between each original pixel and the pixel to be interpolated to obtain an influence factor of each original pixel to the pixel to be interpolated: ω = , \ ≤ m ≤ M

(^ +r ₂ +··· + 3⁄4)

Where ^ is the influence factor of the mth original pixel point on the pixel to be interpolated, and r _m is the degree of correlation between the mth original pixel point and the pixel to be interpolated.

Optionally, calculating, according to an original color value of M original pixel points around the pixel to be interpolated and an influence factor of the M original pixel points to the pixel to be interpolated, calculating a color value of the pixel to be interpolated The steps include:

计算 Calculate the color value of each pixel to be interpolated using the following formula:

R d _p = Re^ x + Ret/ ₂ x +... + Ret/ x ω _Μ

Green _p = Green _x Χ ό) _λ -\ - Green Χ ύ) ₂ -\ H Green _M X (ύ _Μ

Blue _p = Blue x + Blue ₂ x +... + Blue _M x ω _Μ

Where ^, ^^ and ^^ respectively represent the red, green and blue colors of the pixel to be interpolated Component size, ^Ke is the red component of the mth original pixel, ^ is the green component of the mth original pixel, is the blue component of the mth original pixel, and ^ is the mth original pixel pair The influence factor of the pixel to be interpolated, where lm M.

Optionally, the M=4.

A mobile terminal image processing apparatus includes a region of interest detection module and a resolution enhancement module, wherein:

The region of interest detection module is configured to: perform region of interest detection on the image to be processed according to the a priori information of the region of interest obtained in advance;

The resolution enhancement module is configured to: perform resolution improvement on the region of interest detected by the region of interest detection module.

Optionally, the device further includes a prior information obtaining module, where:

The a priori information obtaining module is configured to: obtain the a priori information of the region of interest in advance, where the a priori information acquiring module includes a prompting unit, a first calculating unit, and a saving unit, where:

The prompting unit is configured to: provide one or more images to the user, each image containing one or more targets, prompting the user to calibrate the target of interest in the provided image;

The first calculating unit is configured to: calculate a feature parameter of the user-calibrated target of interest; the saving unit is configured to: save the feature parameter of the target of the user's interest as a priori information to the current target target library of the current user.

Optionally, the region of interest detection module includes a region dividing unit, a second calculating unit, and a comparing unit, where:

The area dividing unit is configured to: divide the image to be processed into one or more object areas according to similar features inside the image;

The second calculating unit is configured to: calculate a feature parameter of the object area;

The comparison unit is configured to: compare a feature parameter of the object area with a feature parameter in a standard interest area target library, and determine an object area with a similarity of the standard feature parameter to a preset threshold as a user Area of interest. Optionally, the resolution enhancement module includes an amplifying unit and a color value calculating unit, where: the amplifying unit is configured to: map the detected region of interest into the amplifying space according to the preset image magnification N The magnification space is N times the region of interest;

The color value calculation unit is configured to: calculate, according to the original pixel point, the color value of the pixel other than the original pixel in the enlarged space according to the original pixel point, and complete the image local resolution improvement.

Optionally, the resolution promotion module further includes a regularization processing unit, where:

The regularization processing unit is configured to: perform regularization processing on the region of interest detected by the region of interest detection module.

Optionally, the color value calculation unit is configured to calculate, according to the original pixel point in the enlargement space, the color value of the pixel other than the original pixel point in the enlarged space according to the original pixel point: The other pixel points outside the pixel point are pixels to be interpolated, and the color value calculation unit performs the following processing on each pixel to be interpolated:

Optionally, the color value calculation unit is configured to calculate a correlation degree between the pixel points to be interpolated and the M original pixel points around the same, and generate an influence factor of the M original pixel points to the pixel to be interpolated:

The color value calculation unit calculates the degree of correlation between each original pixel point and the pixel to be interpolated by using the following formula:

Wherein, indicating the degree of correlation between the mth original pixel point and the pixel to be interpolated, wherein

K m < M , i, j are the horizontal and vertical coordinates of the mth original pixel, respectively, and x and y are the horizontal and vertical coordinates of the pixel to be interpolated, respectively; The color value calculation unit performs normalization processing on the correlation degree of each original pixel point and the pixel to be interpolated to obtain an influence factor of each original pixel point to the pixel to be interpolated: ω = , \ ≤ m ≤ M

(^+r ₂ +··· + 3⁄4)

Optionally, the color value calculation unit is configured to calculate the original color value of the M original pixel points around the pixel to be interpolated and the influence factor of the M original pixel points to the pixel to be interpolated according to the following manner. The color value of the interpolated pixel is mentioned:

The color value calculation unit calculates the color value of each pixel to be interpolated by the following formula: Red _p = Red _l xa) _l +Ret ₂ x<3⁄4 + --- + Red _M χω _Μ

Green _p = Green _x X(0 ₁ + Green ₂ X <3⁄4 HH Green _M X C0 _M

Blue _p = Blue _l Χύ) _ι + Blue ₂ χω ₂ +— h Blue _M X ω _Μ

Where ^, G^"^^te _p respectively represent the red, green, and blue color components of the pixel to be interpolated, ^Re is the red component of the mth original pixel, and ^ is the green of the mth original pixel The component is the blue component of the mth original pixel, and ^ is the influence factor of the mth original pixel on the pixel to be interpolated, where lm M.

Optionally, the M=4.

Compared with the related art, the embodiment of the present invention includes two parts: an adaptive region of interest detection and a resolution enhancement, and the region of interest detection portion fully considers the a priori information of the user's target of interest that is known in advance, thereby improving the detection of the region of interest. The accuracy, the method is more personalized, to meet the needs of different users for different views of the same image; the resolution of the region of interest is improved by only increasing the resolution of the detected region of interest, so The resolution enhancement method in the embodiment of the present invention fully considers the influence of the original pixel point on the pixel to be interpolated, introduces the influence factor of the pixel point of the original image to be interpolated, and calculates the color value of the pixel to be interpolated according to the influence factor, thereby improving Interpolation performance.

The embodiment of the invention provides a function independent and efficient image processing method for the mobile phone user. It is convenient for users to operate, and it is possible for the user to mine more detailed information in the region of interest in the image, and to increase the user experience pleasure.

BRIEF abstract

1 is a flow chart of a method according to Embodiment 1 of the present invention;

2 is a schematic structural view of a device according to Embodiment 2 of the present invention;

3 is a flow chart of an adaptive image region of interest detection and resolution enhancement method according to an application example of the present invention;

4 is a schematic diagram of a simulated region of interest according to an application example of the present invention;

5 is a schematic diagram of an analog image of a region of interest after being enlarged to a designated enlargement space according to an application example of the present invention;

Fig. 6 is a diagram showing the effect of simulating the original image pixel points to be interpolated pixel points in an application example of the present invention. Preferred embodiment of the invention

Considering that in most cases, the user is not interested in the entire image, if the resolution enhancement method is applied to the entire image, the computational complexity will be greatly increased, and the processing time will be greatly increased. In order to ensure better interpolation performance under the premise of low computational complexity, this paper considers only improving the resolution of the region of interest of the user, that is, the resolution of the region of interest of the image is improved. Since the subjective interest information of the user is fully considered, the obtained result is more in line with the user's feelings, and the effect is better.

When detecting the region of interest, considering the user participation information, since the different user regions of interest are often different, the processing results for the same image of different users may be different, more targeted, conforming to user emotions, and the effect is better. In addition, only the image resolution of the user's region of interest in the image is improved, which helps to improve the processing speed, and also satisfies the user's need for more detailed information acquisition of the image region of interest.

In order to make the objects, the technical solutions and the advantages of the present invention more clearly, the embodiments of the present invention will be described in detail below with reference to the accompanying drawings. It should be noted that in the absence of conflict, this application The features of the embodiments and the embodiments may be arbitrarily combined with each other.

Example 1

The image processing method of this embodiment is as shown in FIG. 1 and includes the following steps:

Step 10: performing a region of interest detection on the image to be processed according to the a priori information of the region of interest obtained in advance;

Optionally, the a priori information of the above-mentioned region of interest may be obtained in advance by: providing one or more images to the user, each image containing one or more targets, prompting the user to calibrate himself in the provided image The target of interest; calculating the feature parameter of the user-targeted target of interest; storing the feature parameter of the target of the user's interest as a priori information to the current target target library of the user.

Optionally, the step of performing the region of interest detection on the image to be processed according to the previously obtained region of interest a priori information includes steps 101-103:

Step 101: Divide the image to be processed into one or more object regions according to similar features in the image;

Step 102: Calculate a feature parameter of the object area.

Step 103: Perform similarity comparison between the feature parameter of the target area and the feature parameter in the standard interest area target library, and determine an object area with a similarity of the standard feature parameter greater than a preset threshold as the user interested area;

Step 20: performing resolution improvement on the detected region of interest;

The resolution of the region of interest can be improved by using a known resolution enhancement method. Alternatively, the detected region of interest can be improved by the following method:

Step 201: Map the detected region of interest into the amplifying space according to the preset image magnification N, where the zooming space is N times of the region of interest;

Optionally, if the region of interest is an irregular graphic, for convenience of subsequent processing, the region of interest is first regularized before mapping. The so-called regularization process refers to: in the image to be processed, the following processing is performed on the detected irregular object, as a horizontal line passing through the topmost boundary point and the lowest end boundary point of the object, and the leftmost boundary point of the object and The vertical line of the rightmost boundary point, the closed area divided by the above four lines after the intersection of two and two is a regularized interest Area. A simple processing method is: the original irregular region of interest is circled by a rectangular frame, and the processing of the irregular pattern can be converted into processing of the rectangular region;

Step 202: Calculate, according to the original pixel point in the magnifying space, the color value of the pixel other than the original pixel in the magnifying space, and complete the local resolution of the image; and set the zooming space except the original pixel The other pixels are pixels to be interpolated, and the following processing is performed for each pixel to be interpolated:

Step 2021: Calculate a correlation degree between the pixel to be interpolated and the M original pixel points around the pixel, and generate an influence factor of the M original pixel points to the pixel to be interpolated;

Optionally, the degree of correlation between each original pixel point and the pixel to be interpolated may be calculated by using the following formula:

(^ + r ₂ +··· + 3⁄4)

Where ^ is the influence factor of the mth original pixel point on the pixel to be interpolated, and r _m is the degree of correlation between the mth original pixel point and the pixel to be interpolated;

Step 2022: Calculate a color value of the pixel to be interpolated according to an original color value of the M original pixel points around the pixel to be interpolated and an influence factor of the M original pixel points to the pixel to be interpolated.

Alternatively, the color value of each pixel to be interpolated can be calculated by the following formula:

Red _p = Re j X + Rei ₂ X <» ₂ + · · · + Re _M X <3⁄4

Green _p = Green _x X (0 ₁ + Green ₂ X <3⁄4 HH Green _M X C0 _M

Blue _p = Blue _l Χ ύ) _ι + Blue ₂ χ ω ₂ +— h Blue _M X ω _Μ

Where ^, G^"^^te _p respectively represent the red, green, and blue colors of the pixel to be interpolated Component size, ^Ke is the red component of the mth original pixel, ^ is the green component of the mth original pixel, is the blue component of the mth original pixel, and ^ is the mth original pixel pair The influence factor of the pixel to be interpolated, where lm M.

In the above steps 2021 and 2022, the more the number of original pixel points participating in the calculation, the better the resolution improvement effect, but the higher the computational complexity, so in order to obtain the compromise effect, M=4 is preferable.

Although there are many resolution enhancement methods, in order to obtain a better resolution improvement effect, it often takes a huge computational complexity. The calculation complexity of this method is low, and the low-pass is similar. The filtering method better preserves the global information of the image.

Example 2

The apparatus for implementing the above method is as shown in FIG. 2, and includes a region of interest detection module 20 and a resolution enhancement module 21, wherein:

Optionally, the region of interest detection module 20 includes a region dividing unit 201, a second computing unit 202, and a comparing unit 203, where:

The area dividing unit 201 is configured to: divide the image to be processed into one or more object areas according to similar features inside the image;

The second calculating unit 202 is configured to: calculate a feature parameter of the object area;

The comparison unit 203 is configured to: compare the feature parameters of the target region with the feature parameters in the standard interest region target library, and determine the object region with the similarity of the standard feature parameters greater than the preset threshold as the user sense Area of interest.

The resolution enhancement module 21 is configured to: perform resolution improvement on the region of interest detected by the region of interest detection module.

Optionally, the resolution enhancement module 21 includes an amplification unit 211 and a color value calculation unit 212, where: The amplifying unit 211 is configured to: map the detected region of interest into an amplifying space according to a preset image magnification N, the zooming space being N times of the region of interest;

The color value calculation unit 212 is configured to: calculate, according to the original pixel point, the color value of the pixel other than the original pixel in the enlarged space according to the original pixel point, and complete the image local resolution improvement.

Optionally, the resolution enhancement module 21 may further include a regularization processing unit 213, configured to: perform regularization processing on the region of interest detected by the region of interest detection module.

The color value calculation unit 212 calculates the color value of the pixels other than the original pixel in the amplification space according to the original pixel point in the amplification space according to the original pixel point, and is not described here.

Optionally, the device further includes a prior information obtaining module 22, configured to: obtain the prior information of the region of interest in advance, and include a prompting unit 221, a first calculating unit 222, and a saving unit 223, where:

The prompting unit 221 is configured to: provide one or more images to the user, each image containing one or more targets, prompting the user to calibrate the target of interest in the provided image;

The first calculating unit 222 is configured to: calculate a feature parameter of the user-calibrated target of interest; the saving unit 223 is configured to: save the feature parameter of the target of interest of the user as a priori information to the standard target target library of the current user.

Application example

自适应Adaptive image region of interest resolution enhancement method can better achieve the local zooming effect of the image region of interest. The following is an example of adaptive image region detection and region of interest resolution enhancement method with M=4 as an example. Description. As shown in Figure 3.

First, the image region of interest detection:

Before performing image area detection, you need to use the following steps to establish a target library of interest to the user to obtain a priori information:

Step 11: Provide multiple images containing multiple targets; Step 12: prompting the user to separately calibrate the target of interest in the plurality of images; Step 13: calculating a feature parameter of the target of interest of the user calibration;

Step 14: Statistics the target of interest to the user, and save the feature parameters of the target of interest to the current target target library of the current user.

Through the above steps, the establishment of the standard data of the target object of interest of the user is completed, and is used for determining the region of interest of the user in the subsequent image processing. The feature parameters of the above-mentioned target of interest to the user may be referred to as prior information. In order to adapt to changes in user preferences, the above steps can be performed periodically to update the user's target library of interest.

The image region detection process includes:

Step 21: Divide the input image to be processed into different object regions according to similar features in the image;

The similar features inside the image refer to the similar similarity between the inside of the image, that is, the pixel points with the physical position are closer, and the color values are similar. Therefore, the object area can be divided according to similar features inside the image, i.e., color values.

Step 22: Calculate characteristic parameters of each object area mentioned above;

Step 23: Comparing the feature parameters of each of the object regions with the feature parameters in the target region of the standard region of interest, and the object region whose similarity with the standard feature parameter is greater than the preset threshold T is the region of interest of the user.

Through the above steps, the adaptive region of interest detection process of the image to be processed is completed. In order to enhance the information of the target and improve the resolution, the resolution of the extracted region of interest can be improved.

Second, the image adaptive region of interest resolution is improved:

Step 31: Map the region of interest to the zoom space according to the image magnification N of the preset (user setting or default setting), where the zoom space is N times of the region of interest; wherein N is a controllable parameter, the user You can enter the determination in advance.

Optionally, before performing the enlargement, for the convenience of subsequent processing, the region of interest may be firstly regularized, that is, processed into a regular graph. As shown in FIG. 4, A is a region of interest of the user, and after regularization, Obtaining B, the rule pattern is preferably square or rectangular.

The enlargement process is as shown in Fig. 5. The left side of Fig. 5 shows the position of four pixel points in the original image before enlargement. After zooming in four times, the position of the four pixel points is as shown in the right figure of Fig. 5.

Step 32: Calculate the color value of the pixels other than the original pixel in the magnifying space according to the original pixel in the magnifying space, and complete the local resolution of the image.

The above step 32 specifically includes the following steps:

Let the 4 original pixels be , , , , and . , the pixel to be interpolated within the original pixel point is P;

Step 321 : Calculate the degree of correlation between the pixel to be interpolated and the four original pixel points around the pixel, and generate an influence factor of the four original pixel points to the pixel to be interpolated;

Taking the P point shown in Figure 5 as an example, calculate the correlation degree of the four original pixel points A to B, respectively, and normalize the correlation degree to obtain the color value of the four original pixel points and the color value of the P point. The impact factor, " ₂ , " ₃ and " ₄ , as shown in Figure 6. According to Figure 6, set the coordinates of point A ( ^z ', , then the coordinates of point B ('·, ^·· + Λ , C point coordinates ( ^{+ ΛΓ} ' ), D point coordinates ( ^{+ Λ} ^· ^{+ ΛΓ} ), set to be interpolated The pixel point Ρ coordinate is ( ^χ , , first calculate the correlation between the four points of B, C, D and the pixel point P to be interpolated as follows, the following negative index Euclidean distance method can be used:

Where ^ ^ denotes the degree of correlation between the defect and the pixel to be interpolated, and so on, and ^ are respectively:

r _ -((xiW) ² +(j- ) ² ) r _ _p -((xiN†+( _y -jN†~) Then calculate the eight, B, C, D four points to interpolate according to the above correlation degree The influence factor of pixel P is to normalize the above correlation degree as follows:

r_BP

^ i

Among them, the influence factor of point A on point P, and the influence factors of point B, C and D on P are respectively , and . Step 322: Calculate a color value of the pixel to be interpolated according to an original color value of four original pixel points around the pixel to be interpolated and an influence factor of the four original pixel points to the pixel to be interpolated; as shown in FIG. 5 and FIG. 5 For example, the color values of the pixels P to be interpolated are evaluated according to the above-mentioned influence factors and the four-point color values of A, B, C, and D, as follows:

Re d _p = Red _A xo _l +Red _{B 2} +Red _{c 3} +Red _D x ₄

Green _p = Green _A o + Green _B ω ₂ + Green _c x<3⁄4 + Green _D x ω _Λ

Blue _p = Blue _A χω _ι + Blue _B ω ₂ + Blue _c ω ₃ + Blue _D x ω _Λ

Where ^, ^ and 53⁄4^ respectively represent the red, green and blue color components of the pixel p to be interpolated.

After the above-mentioned image adaptive region of interest detection and resolution enhancement step, the image region of interest is adaptively determined, and the resolution is improved according to the user's needs, thereby improving the information amount of the image of the region of interest, and the user can view more effectively. Image detail information, the process ends.

One of ordinary skill in the art will appreciate that all or a portion of the above steps may be accomplished by a program instructing the associated hardware, such as a read-only memory, a magnetic disk, or an optical disk. Alternatively, all or part of the steps of the above embodiments may also be implemented using one or more integrated circuits. Correspondingly, each module/unit in the above embodiment may be implemented in the form of hardware or in the form of a software function module. The invention is not limited to any specific form of combination of hardware and software.

It is a matter of course that the invention may be embodied in various other forms and modifications without departing from the spirit and scope of the invention. Industrial applicability

Compared with the related art, the embodiment of the present invention includes two parts: an adaptive region of interest detection and a resolution enhancement, and the region of interest detection portion fully considers the a priori information of the user's target of interest that is known in advance, thereby improving the detection of the region of interest. The accuracy, the method is more personalized, to meet the needs of different users for different views of the same image; the resolution of the region of interest is improved by only increasing the resolution of the detected region of interest, so The resolution enhancement method in the embodiment of the present invention fully considers the influence of the original pixel point on the pixel to be interpolated, introduces the influence factor of the pixel point of the original image to be interpolated, and calculates the color value of the pixel to be interpolated according to the influence factor, thereby improving Interpolation performance. The embodiment of the invention provides a unique and efficient image processing method for the mobile phone user, which is convenient for the user to operate, and provides the user with the possibility of mining more detailed information in the region of interest in the image, thereby increasing the user experience pleasure. Therefore, the present invention has strong industrial applicability.

Claims

Claim

1. A mobile terminal image processing method, comprising:

The region of interest is detected according to the a priori information of the region of interest obtained in advance; the resolution of the detected region of interest is improved.

The mobile terminal image processing method according to claim 1, further comprising: pre-obtaining in the following manner before the step of performing the region of interest detection on the image to be processed according to the previously obtained region of interest prior information; The prior information of the region of interest:

Calculating characteristic parameters of the user-targeted target of interest;

The feature parameters of the target of interest of the user are saved as a priori information to the current user's standard interest object library.

The mobile terminal image processing method according to claim 2, wherein the step of performing region of interest detection on the image to be processed according to the previously obtained region of interest prior information comprises: processing the image to be processed according to similar features within the image Dividing into one or more object regions; calculating feature parameters of the object region;

The mobile terminal image processing method according to claim 1 or 2 or 3, wherein the step of performing resolution improvement on the detected region of interest comprises:

Mapping the detected region of interest into an amplification space according to a preset image magnification N, the magnification space being N times the region of interest;

The mobile terminal image processing method according to claim 4, wherein, before the step of mapping the detected region of interest to the enlarged space according to the preset image magnification N, The method further includes:

The region of interest is regularized.

The mobile terminal image processing method according to claim 4, wherein the step of calculating, according to the original pixel point in the enlargement space, the color value of the pixel other than the original pixel in the enlarged space by using the interpolation method comprises: :

Let other pixels in the magnifying space besides the original pixel point be the pixels to be interpolated, and perform the following processing on each pixel to be interpolated:

The mobile terminal image processing method according to claim 6, wherein the calculating the degree of correlation between the pixel to be interpolated and the M original pixel points around the M pixel, and generating the M original pixel points to the pixel to be interpolated The steps of the impact factor include:

(^ +r ₂ +··· + 3⁄4)

The mobile terminal image processing method according to claim 7, wherein the original color value of the M original pixel points around the pixel to be interpolated and the M original pixel points are to the to-be-interpolated The interpolation factor of the interpolation pixel, the step of calculating the color value of the pixel to be interpolated includes: 计算 calculating the color value of each pixel to be interpolated by using the following formula:

R d _p = Re^ x + Ret/ ₂ x +... + Ret/ x ω _Μ

Green _p = Green _x Χ ό) _λ -\ - Green Χ ύ) ₂ -\ H Green _M X (ύ _Μ

Blue _p = Blue x + Blue ₂ x +... + Blue _M x ω _Μ

Where ^, ^^ and ^^ respectively represent the red, green and blue color components of the pixel to be interpolated, ^Re is the red component of the mth original pixel, and ^ is the green component of the mth original pixel. For the blue component of the mth original pixel point, ^ is the influence factor of the mth original pixel point on the pixel to be interpolated, where lm M.

The mobile terminal image processing method according to claim 6 or 7 or 8, wherein

M=4.

10. A mobile terminal image processing apparatus, comprising: a region of interest detection module and a resolution enhancement module, wherein: the region of interest detection module is configured to: be interested in processing an image according to a prior information of a prior region of interest Area detection

The mobile terminal image processing apparatus according to claim 10, further comprising a priori information acquisition module, wherein:

The a priori information obtaining module is configured to: obtain the a priori information of the region of interest in advance, where the a priori information acquiring module includes a prompting unit, a first calculating unit, and a saving unit, where: the prompting unit is configured to: The user provides one or more images, each image containing one or more targets, prompting the user to calibrate the target of interest in the provided image; the first computing unit is configured to: calculate the user's calibrated interest The feature parameter of the target; the saving unit is configured to: save the feature parameter of the target of interest of the user as a priori information to the standard target object library of the current user.

The image processing apparatus for a mobile terminal according to claim 11, wherein said region of interest The domain detection module includes a region division unit, a second calculation unit, and a comparison unit, where:

The comparison unit is configured to: compare a feature parameter of the object area with a feature parameter in a standard interest area target library, and determine an object area with a similarity of the standard feature parameter to a preset threshold as a user Area of interest.

The apparatus according to claim 10 or 11 or 12, wherein the resolution enhancement module comprises an amplification unit and a color value calculation unit, wherein:

The amplifying unit is configured to: map the detected interest area into the enlargement space according to a preset image magnification N, wherein the enlargement space is N times of the region of interest;

The apparatus of claim 13, wherein the resolution enhancement module further comprises a regularization processing unit, wherein:

The device according to claim 13, wherein the color value calculation unit is configured to calculate, in the enlargement space, the pixel points other than the original pixel point in the enlargement space according to the original pixel point in the enlargement space. Color value:

It is assumed that other pixels in the enlargement space except the original pixel point are pixels to be interpolated, and the color value calculation unit performs the following processing on each pixel to be interpolated:

Calculating a color value of the pixel to be interpolated according to an original color value of the M original pixel points around the pixel to be interpolated and an influence factor of the M original pixel points to the pixel to be interpolated.

The apparatus according to claim 15, wherein the color value calculation unit is configured to calculate a degree of correlation between the pixel to be interpolated and the M original pixel points around the same, and generate the M original pixel points to the location The influence factor of the interpolated pixel is mentioned:

The color value calculation unit performs normalization processing on the correlation degree of each original pixel point and the pixel to be interpolated to obtain an influence factor of each original pixel point to the pixel to be interpolated: ω = , \ ≤ m ≤ M

(^ +r ₂ +··· + 3⁄4) where ^ is the influence factor of the mth original pixel point on the pixel to be interpolated, r _m is the mth original pixel point and the pixel to be interpolated Relevance.

The apparatus according to claim 16, wherein the color value calculation unit is configured to: according to the original color value of the M original pixel points around the pixel to be interpolated and the M original pixel points to the waiting Interpolating the influence factor of the pixel, and calculating the color value of the pixel to be interpolated:

The color value calculation unit calculates the color value of each pixel to be interpolated by the following formula: Red _p = Re j X + Rei ₂ X <» ₂ + · · · + Re _M X <3⁄4

Green _p = Green _x X (0 ₁ + Green ₂ X <3⁄4 HH Green _M X C0 _M

Blue _p = Blue _l Χ ύ) _ι + Blue ₂ χ ω ₂ +— h Blue _M X ω _Μ

18. Apparatus according to claim 15 or 16 or 17, wherein said M = 4.