Multi-factor driven micro map image visual balance degree calculation method

Abstract

The invention discloses a multi-factor-driven micro map image visual balance degree calculation method, and belongs to the field of map cartography and artificial intelligence. The method comprises the following steps: firstly, performing morphological reconstruction and membership matrix division on each micro map image to optimize the distribution characteristics of the image and determine the attribution degree of each element, so as to realize the classification of the drawing configuration elements based on robust fuzzy C-means clustering; thirdly, three visual perception factors including brightness, contrast and saliency are introduced, the weight, details and focuses of vision are concerned respectively, and four main factors for visual density quantification are jointly formed in combination with position factors influencing element layout and distribution; and finally, according to the visual density, a We-map visual balance calculation (W-VBC) model of the image of the micro map is constructed, and according to the visual density, the We-map visual balance calculation (W-VBC) model of the image of the micro map is constructed. According to the method, for the micro-map image with popular and personalized characteristics in the we-media era, the visual perception of the micro-map image is reflected more truly through image element classification and multi-factor quantification, and the problem of uncertainty existing in artificial subjective judgment of the visual balance degree of the map is effectively solved.

Description

Technical Field

[0001] The present invention belongs to the field of cartography and artificial intelligence, and specifically is a multi-factor driven method for calculating the visual balance of a micro-map image. Background Art

[0002] Visual balance originated from the field of art. It is an important research topic in fields such as psychology and psychophysics. It is also an important factor affecting the artistry and scientific nature of maps. It plays a very important role in improving the efficiency of map information transmission and visual beauty.

[0003] In the traditional sense, maps are made by professionally trained technicians. The map elements are fixed and follow certain cartographic specifications. The map elements are basically balanced. However, with the advent of the self-media era, the threshold for map production has quietly lowered, providing unlimited possibilities for popular map creative design and diversified production. This change has given rise to the novel concept of "micro-map". As a grassroots force, it breaks the boundaries of traditional map production with its civilian, personalized, low threshold, micro-content, strong interaction and fast dissemination. Micro-maps do not require high mathematical accuracy. Their creators are spread across all walks of life, and the design level therefore presents a diversified appearance. As a result, this popularization trend also brings challenges: many micro-map images often seem to be unable to achieve visual balance, which not only weakens the effective communication efficiency of geographic information, but also affects the aesthetic value of the map itself as a work of art. There is currently a lack of unified and objective standards for the evaluation of the visual balance of micro-maps. It relies more on the experience accumulated by cartographers in long-term practice and their personal aesthetic cultivation. This undoubtedly increases the subjective color of the evaluation process, making how to maintain content innovation while improving visual beauty a topic worthy of in-depth discussion.

[0004] Through the summary of the prior art, it is found that there are at least the following problems:

[0005] 1) Existing methods mostly use manually annotated polygon box selection to extract label data of the main map and auxiliary map elements for discrimination, which is not applicable to maps such as micro maps that mainly use personalized annotation elements and have no standard map elements.

[0006] 2) The existing visual balance calculation methods have problems such as simple and superficial selection of influencing factors, lack of specificity, and inaccurate calculations, and are unable to fully focus on key factors such as visual weight, details, and focus.

[0007] 3) Existing visual balance models for map images often face unreasonable construction problems, such as complex parameters and improper weight distribution, which lead to strong subjectivity and limited accuracy in the judgment results. These problems seriously restrict the scientific evaluation and improvement of the visual balance of map images. Therefore, it is urgent to develop a more reasonable and comprehensive model to accurately quantify the visual balance of map images.

[0008] Therefore, how to provide a method that comprehensively considers aspects such as label data processing, visual density factor selection, and construction of a reasonable model is an issue that technicians in this field urgently need to solve. Summary of the invention

[0009] In view of this, the present invention aims at the shortcomings of the above-mentioned methods and provides a new method with high application value, which is simple and feasible and is used to quantitatively measure the visual balance quality of micromap images. It effectively solves the problems of the existing methods, such as the low versatility of the map element label data processing method, the lack of targetedness in the selection of visual density influencing factors, and the imperfect construction of the visual balance model.

[0010] In order to achieve the above object, the present invention adopts the following technical solution:

[0011] A multi-factor driven method for calculating visual balance of a micro map image, comprising:

[0012] S1: Classify the various configuration elements of the micromap image through robust fuzzy C-means clustering, and determine the visual center of each configuration element;

[0013] S2: Quantify the brightness, contrast, saliency and position factors of the layout elements, and calculate the visual density of each layout element;

[0014] S3: According to the visual center and visual density of the drawing configuration elements, the W-VBC model is constructed based on the moment balance principle.

[0015] Preferably, in S1, the classification method of the micro map image surface configuration elements is as follows:

[0016] S21: Morphological reconstruction through erosion and dilation operations;

[0017] S22: Perform membership matrix division on the morphologically reconstructed image.

[0018] Preferably, in S2, the visual density quantification method of each drawing configuration element is as follows:

[0019] S31: Calculate the average brightness, Lab color space contrast and frequency domain multi-scale difference visual significance of the map configuration elements;

[0020] S32: Partition the elements of each drawing according to the Rudolf Arnheim structural network diagram, and quantify the position factor by the belonging distance.

[0021] Preferably, in S3, the W-VBC model construction method is as follows:

[0022] S41: Divide the mini map image into four regions: upper left, upper right, lower left, and lower right;

[0023] S42: Obtain the visual weight of each area according to the visual density and the area of ​​the elements configured on the drawing, and multiply it by the visual gravity moment to obtain the visual gravity distance;

[0024] S43: Calculate overall, up-down, and left-right balance indices by combining the visual gravity moments of the four regions.

[0025] Compared with the prior art, the present invention has the following beneficial technical effects:

[0026] Accurate classification of map elements: Through morphological reconstruction and membership matrix partitioning, combined with robust fuzzy C-means clustering technology, the present invention achieves accurate classification of map configuration elements in micro-map images. This method not only optimizes the distribution characteristics of the image, but also clarifies the degree of belonging of each element, providing a solid foundation for the subsequent calculation of visual balance.

[0027] Multi-factor comprehensive quantification of visual perception: The three key visual perception factors of brightness, contrast, and saliency, as well as the position factor, are introduced to jointly construct a four-dimensional model for visual density quantification. This multi-factor driven method can comprehensively and meticulously capture the visual features in micro-map images, including visual weight, detail levels and focal areas, as well as the spatial layout of elements, thereby more realistically reflecting the visual perception of the image.

[0028] Scientific visual balance calculation model: Based on the above multi-factor quantification results, the W-VBC model was constructed to calculate the visual balance of the micro-map image. This model effectively reduces the uncertainty caused by human subjective judgment and improves the objectivity and accuracy of visual balance evaluation. BRIEF DESCRIPTION OF THE DRAWINGS

[0029] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the prior art descriptions are briefly introduced below. Obviously, the drawings described below are only embodiments of the present invention, and for ordinary technicians in this field, other drawings can be obtained based on the provided drawings without creative work.

[0030] Figure 1 Flowchart of the method proposed for the present invention

[0031] Figure 2 Quantized partitioning example diagram for position factor

[0032] Figure 3 Example diagram of the partition of the micro map image surface elements DETAILED DESCRIPTION

[0033] The technical scheme in the embodiment of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiment of the present invention. Obviously, the described embodiment is only a part of the embodiment of the present invention, not all of the embodiments. Based on the embodiment of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of the present invention.

[0034] See attached Figure 1 The embodiment of the present invention discloses a method for calculating the visual balance of a micro-map image driven by multiple factors, comprising the following steps:

[0035] Step 1: Classify the configuration elements of each surface of the micromap image.

[0036] Step 1: Perform dilation-based reconstruction on the micro-map images in sequence and corrosion reconstruction The morphological closed reconstruction optimization of where δ(·) and ε(·) represent dilation and erosion operations, respectively.

[0037] Step 2: Reconstruct the morphological image f′=R C The grayscale histogram of (f) is clustered, and the optimization process can be expressed as where u kl represents the fuzzy membership of the gray value l relative to cluster k, v k is the prototype value of the kth cluster, Represents the number of pixels in the image, f l ′ is the gray level of f′, q is the number of gray levels contained in f′, and c is the number of clusters.

[0038] Step 3: Solve the clustering optimization function to obtain and And iterate it until max{U (t) -U (t+1)}<η, and obtain a stable membership partition matrix U=[u kl ] c×N , and finally the membership partition matrix is ​​corrected using the median filter, where η is the minimum error threshold.

[0039] Step 2: Quantification of factors affecting the visual density of drawing configuration elements.

[0040] Step 1: Obtain the brightness value of each drawing configuration element through Lab color transformation, and then quantify the brightness factor according to the average brightness of each drawing configuration element.

[0041] Step 2: Contrast quantification is performed in the Lab color space, and contrast information is calculated based on the Euclidean distance between the average Lab value of each image configuration element and the average Lab value of all other image configuration elements.

[0042] Step 3: First, calculate the corresponding characteristic spectrum by weighted integration of the differential Gaussian features in the frequency domain in is the Fourier transform of the i-th image configuration element, is the Fourier spectrum of the Gaussian kernel with standard deviation σ, With G σ Similarly, w σ is the weight of σ. Then, by performing scale filtering on the saliency map of each channel, salient objects within a specific scale are extracted. where s i,k is the feature of the kth channel, ψ(·) is the scale filter, M i,k is the characteristic spectrum of the kth channel obtained by, Represents the inverse Fourier transform. Finally, the saliency feature map is integrated by weighting and summing the channel saliency according to Hoyer sparseness Among them, Sparse(s i,k ) is s i,k Hoyer sparsity.

[0043] Step 4: First, combine the Rudolf Arnheim structural network diagram to divide the map into 5 areas, see the attached Figure 2 Then, the regional affiliation of the drawing configuration element is determined according to the distance between the visual center of the drawing configuration element and the center of each region; finally, a weighted calculation is performed on the drawing configuration element according to the distance between the visual center of the drawing configuration element and the center of the region to which it belongs.

[0044] Step 3: W-VBC model construction.

[0045] Step 1: Divide the micromap image into four areas: upper left, upper right, lower left, and lower right. Figure 3 , the visual gravity moment of each area is the sum of the visual gravity distances of all the surface configuration elements contained in the area, that is In the formula, S i , i are the area and visual density of the i=1, 2,…, nth drawing elements, L iis the distance from the visual center of the ith element to the visual center of the entire map, and n is the number of elements in the θ=1, 2, 3, 4 regions.

[0046] Step 2: Calculate the overall balance index of the map image visual balance based on the visual gravity moments VM1, VM2, VM3, and VM4 of the upper left, upper right, lower left, and lower right regions. Up and down balance index C TB =(VM T -VM B ) / max(VM T ,VM B ) and left-right balance index C LR =(VM L -VM R ) / max(VM L ,VM R ), where VM T =VM1+VM2, VM B =VM3+VM4, VM L =VM1+VM3,VM R =VM2+VM4.

[0047] In this specification, each embodiment is described in a progressive manner, and each embodiment focuses on the differences from other embodiments. The same or similar parts between the embodiments can be referred to each other. For the device disclosed in the embodiment, since it corresponds to the method disclosed in the embodiment, the description is relatively simple, and the relevant parts can be referred to the method part.

[0048] The above description of the disclosed embodiments enables one skilled in the art to implement or use the present invention. Various modifications to these embodiments will be apparent to one skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present invention. Therefore, the present invention will not be limited to the embodiments shown herein, but rather to the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A multi-factor driven method for calculating visual balance of micro-map images, characterized in that: include: S1: Classify the various configuration elements of the micromap image through robust fuzzy C-means clustering and determine the visual center of each configuration element; S2: Quantify the brightness, contrast, saliency and position factors of the layout elements, and calculate the visual density of each layout element; S3: According to the visual center and visual density of the drawing configuration elements, the W-VBC model is constructed based on the moment balance principle.

2. The multi-factor driven micro map image visual balance calculation method according to claim 1 is characterized in that: In S1, the classification method of the micro map image surface configuration elements is as follows: S21: Morphological reconstruction through erosion and dilation operations; S22: Perform membership matrix division on the morphologically reconstructed image.

3. The multi-factor driven micro map image visual balance calculation method according to claim 1 is characterized in that: In S2, the visual density quantification method of each drawing configuration element is as follows: S31: Calculate the average brightness, Lab color space contrast and frequency domain multi-scale difference visual significance of the map configuration elements; S32: Partition the elements of each drawing according to the Rudolf Arnheim structural network diagram, and quantify the position factor by the belonging distance.

4. The multi-factor driven micro map image visual balance calculation method according to claim 1 is characterized in that: In S3, the W-VBC model construction method is as follows: S41: Divide the mini map image into four regions: upper left, upper right, lower left, and lower right; S42: Obtain the visual weight of each area according to the visual density and the area of ​​the elements configured on the drawing, and multiply it by the visual gravity moment to obtain the visual gravity distance; S43: Calculate overall, up-down, and left-right balance indices by combining the visual gravity moments of the four regions.

Images