WO2023148411A1 - Method for calibrating graphs - Google Patents
Method for calibrating graphs Download PDFInfo
- Publication number
- WO2023148411A1 WO2023148411A1 PCT/ES2022/070052 ES2022070052W WO2023148411A1 WO 2023148411 A1 WO2023148411 A1 WO 2023148411A1 ES 2022070052 W ES2022070052 W ES 2022070052W WO 2023148411 A1 WO2023148411 A1 WO 2023148411A1
- Authority
- WO
- WIPO (PCT)
- Prior art keywords
- graph
- graphs
- units
- axes
- artificial intelligence
- Prior art date
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/20—Information retrieval; Database structures therefor; File system structures therefor of structured data, e.g. relational data
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F7/00—Methods or arrangements for processing data by operating upon the order or content of the data handled
Definitions
- the present application refers to a graph calibration procedure that allows transforming a series of different graphs to allow their easy comparison automatically.
- Any series of points can be graphed in an infinite number of ways, making it difficult to compare with other series of points.
- the cause is that the scale, or distance between the divisions set on the axes, is assigned independently for each axis.
- the invention consists of a chart calibration procedure according to the first claim. Its different variants solve the problems outlined.
- the chart calibration procedure allows you to match the axes of two or more charts. It is done using artificial intelligence and comprises several stages:
- It can comprise a previous stage of localizing graphics in one or more documents, also using artificial intelligence.
- Figure 1 Schematic example of realization.
- the procedure is described from a case in which two graphs (1,2) of one or more documents wish to be compared.
- Graphs (1,2) represent three series of points.
- the user can select them or they can be detected by the artificial intelligence process that performs the calibration, since the graphics have a standardized format, with at least two orthogonal bars and a series of designs in the central part.
- the procedure begins by detecting the units of both axes by character recognition and using artificial intelligence.
- the units are not comparable, it is assessed whether there can be a transformation. For example, change of units from the imperial system to the metric system (pounds to kilograms), to the same multiple (KPa to Pa), from different units (calories to Joules), from absolute to relative values (%), etc.
- a series of priorities can be defined, for example, giving priority to relative scales, to the international system... It will also detect if there is an inversion of the variables, that is, a variable is on the abscissa in a graph and on the ordinate on the other or if the scale increases from right to left on one axis.
- the procedure stops. For example, if one graph is "length-time" and the other is "percent compound - reaction yield. If one unit on each graph is compatible, but not the other, percentages (increasing-decreasing) will be used on the other unit of each graph.
- the units are the same, but on scales or ranges so different that any comparison is broken. For example, they can be very distant dates or compare values in kilometers with values in microns. In this second case, a step to relative units (percentage%) may be applicable.
- the values of the variables on the axis are read and stored in memory for each graph.
- the position of each curve on the graph is then estimated, taking into account any confidence intervals. From these data, a final graph (3) is drawn where the corresponding curves are represented.
- the curves of each document can start from different abscissa values, which must be adequately represented, as can be seen in figure 1.
- Each curve can be represented with a different color, type of line.... If the graphs use relative values, they are preferably modified so that the starting value coincides in the same abscissa, applying the corresponding rule of three.
- the final graph may comprise a separate grid, not matching the scale of the corresponding axes, but defined as a proportion of the axis scales in the original graphs.
Abstract
Disclosed is a method for calibrating graphs, by means of artificial intelligence, which coordinates the abscissas and ordinates of two or more graphs to compare them. The method comprises: recognising the units of the abscissa and ordinate axes; comparing and transforming the units of both axes; calculating a series of points of each curve of each graph; and redrawing the curves in a final graph (3).
Description
DESCRIPCIÓN DESCRIPTION
Procedimiento de calibración de gráficos Graphics Calibration Procedure
SECTOR DE LA TÉCNICA TECHNIQUE SECTOR
La presente solicitud se refiere a un procedimiento de calibración de gráficos que permite transformar una serie de gráficos diferentes para permitir su comparación sencilla de forma automática. The present application refers to a graph calibration procedure that allows transforming a series of different graphs to allow their easy comparison automatically.
ESTADO DE LA TÉCNICA STATE OF THE ART
En la literatura científica, económica, política y otros tipos de escritos es frecuente aportar gráficos en los que las abscisas representan una variable y las ordenadas una segunda variable y escala. Al comparar gráficos de un mismo documento o de dos documentos diferentes es difícil realizar la comparación si las variables de cada eje son diferentes (por ejemplo, porcentaje o valor absoluto, fecha de inicio y final, valor de la ordenada en el cruce de ambos ejes...). Este tipo de manipulaciones se realizan a menudo para ocultar o sugerir resultados según el interés del redactor. In scientific, economic, political and other types of writing, it is common to provide graphs in which the abscissas represent one variable and the ordinates represent a second variable and scale. When comparing graphs from the same document or from two different documents, it is difficult to make the comparison if the variables of each axis are different (for example, percentage or absolute value, start and end date, value of the ordinate at the intersection of both axes ...). These types of manipulations are often done to hide or suggest results according to the editor's interest.
Cualquier serie de puntos se puede representar gráficamente de infinitas maneras, lo cual dificulta su comparación con otras series de puntos. La causa es que la escala, o distancia entre las divisiones establecidas en los ejes, se asigna de forma independiente para cada eje. Any series of points can be graphed in an infinite number of ways, making it difficult to compare with other series of points. The cause is that the scale, or distance between the divisions set on the axes, is assigned independently for each axis.
Por lo tanto, es deseable poder realizar comparaciones de forma sencilla, ya sea de gráficos de un único documento o de varios. Therefore, it is desirable to be able to make comparisons easily, whether it be graphs from a single document or several.
El solicitante no conoce ningún sistema que permita resolver todos estos problemas de forma tan sencilla como la invención. The Applicant is not aware of any system that makes it possible to solve all these problems as simply as the invention.
BREVE EXPLICACIÓN DE LA INVENCIÓN
La invención consiste en un procedimiento de calibración de gráficos según la reivindicación primera. Sus diferentes variantes resuelven los problemas reseñados. BRIEF EXPLANATION OF THE INVENTION The invention consists of a chart calibration procedure according to the first claim. Its different variants solve the problems outlined.
El procedimiento de calibración de gráficos permite igualar los ejes de dos o más gráficos. Se realiza mediante inteligencia artificial y comprende varias etapas: The chart calibration procedure allows you to match the axes of two or more charts. It is done using artificial intelligence and comprises several stages:
1. Reconocer las unidades de los ejes de abscisas y ordenadas. 1. Recognize the units of the abscissa and ordinate axes.
2. Comparar y transformar las unidades de ambos ejes. 2. Compare and transform the units of both axes.
3. Calcular una serie de puntos de cada curva de cada gráfico. 3. Calculate a series of points for each curve of each graph.
4. Redibujar las curvas en un gráfico final. 4. Redraw the curves on a final graph.
Puede comprender una etapa previa de localización de gráficos en uno o más documentos, también mediante inteligencia artificial. It can comprise a previous stage of localizing graphics in one or more documents, also using artificial intelligence.
Otras variantes se aprecian en el resto de la memoria. Other variants are appreciated in the rest of the memory.
DESCRIPCIÓN DE LAS FIGURAS DESCRIPTION OF THE FIGURES
Para una mejor comprensión de la invención, se incluyen las siguientes figuras. For a better understanding of the invention, the following figures are included.
Figura 1 : Ejemplo esquemático de realización. Figure 1: Schematic example of realization.
MODOS DE REALIZACIÓN DE LA INVENCIÓN MODES OF CARRYING OUT THE INVENTION
A continuación, se pasa a describir de manera breve un modo de realización de la invención, como ejemplo ilustrativo y no limitativo de ésta. An embodiment of the invention is briefly described below, as an illustrative and non-limiting example thereof.
El procedimiento se describe a partir de un caso en el que dos gráficos (1,2) de uno o más documentos desean ser comparados. Los gráficos (1,2) representan tres series de puntos. El usuario puede seleccionarlos o ser detectados por el propio proceso de inteligencia artificial que realiza la calibración, dado que los gráficos tienen un formato estandarizado, con al menos dos barras ortogonales y una serie de diseños en la parte central.
En ambos casos, el procedimiento se inicia detectando las unidades de ambos ejes por reconocimiento de caracteres y usando la inteligencia artificial. Así, si las unidades no son comparables, se valora si puede haber transformación. Por ejemplo, cambio de unidades del sistema imperial al sistema métrico (libras a kilogramos), al mismo múltiplo (KPa a Pa), de diferentes unidades (calorías a Julios), de valores absolutos a relativos (%), etc. Para ello se puede definir una serie de prioridades, por ejemplo dando prioridad a escalas relativas, al sistema internacional... Se detectará también si hay inversión de las variables, es decir, una variable está en las abscisas en un gráfico y en las ordenadas en el otro o si la escala aumenta de derecha a izquierda en un eje. The procedure is described from a case in which two graphs (1,2) of one or more documents wish to be compared. Graphs (1,2) represent three series of points. The user can select them or they can be detected by the artificial intelligence process that performs the calibration, since the graphics have a standardized format, with at least two orthogonal bars and a series of designs in the central part. In both cases, the procedure begins by detecting the units of both axes by character recognition and using artificial intelligence. Thus, if the units are not comparable, it is assessed whether there can be a transformation. For example, change of units from the imperial system to the metric system (pounds to kilograms), to the same multiple (KPa to Pa), from different units (calories to Joules), from absolute to relative values (%), etc. For this, a series of priorities can be defined, for example, giving priority to relative scales, to the international system... It will also detect if there is an inversion of the variables, that is, a variable is on the abscissa in a graph and on the ordinate on the other or if the scale increases from right to left on one axis.
Si la inteligencia artificial detecta que las unidades no son compatibles, por ejemplo porque ninguna de las unidades de cada gráfico (1,2) está relacionada con el otro, se detiene el procedimiento. Por ejemplo, si un gráfico es "longitud-tiempo" y el otro "porcentaje de compuesto - rendimiento de la reacción. Si una unidad de cada gráfico es compatible, pero no la otra, se utilizarán porcentajes (crecimiento-decrecimiento) en la otra unidad de cada gráfico. If the artificial intelligence detects that the units are not compatible, for example because none of the units in each graph (1,2) are related to the other, the procedure stops. For example, if one graph is "length-time" and the other is "percent compound - reaction yield. If one unit on each graph is compatible, but not the other, percentages (increasing-decreasing) will be used on the other unit of each graph.
También es posible que las unidades sean ¡guales, pero en escalas o rangos tan diferentes que se rompa cualquier comparación. Por ejemplo, pueden ser fechas muy distanciadas o comparar unos valores en kilómetros con unos valores en mieras. En este segundo caso puede ser aplicable un paso a unidades relativas (porcentaje...). It is also possible that the units are the same, but on scales or ranges so different that any comparison is broken. For example, they can be very distant dates or compare values in kilometers with values in microns. In this second case, a step to relative units (percentage...) may be applicable.
Una vez detectadas las unidades, se leen los valores de las variables en el eje y se guardan en memoria, para cada gráfica. A continuación, se estima la posición de cada curva en la gráfica, teniendo en cuenta cualquiera intervalo de confianza. A partir de estos datos, se dibuja un gráfico final (3) donde se representan las correspondientes curvas. Las curvas de cada documento pueden partir de diferentes valores de abscisas, lo cual ha de representarse adecuadamente, como se aprecia en la figura 1. Cada curva se podrá representar con un color, tipo de línea... diferente.
Si los gráficos utilizan valores relativos, preferiblemente se modifican para que el valor de partida coincida en la misma abscisa, aplicando la correspondiente regla de tres. Once the units are detected, the values of the variables on the axis are read and stored in memory for each graph. The position of each curve on the graph is then estimated, taking into account any confidence intervals. From these data, a final graph (3) is drawn where the corresponding curves are represented. The curves of each document can start from different abscissa values, which must be adequately represented, as can be seen in figure 1. Each curve can be represented with a different color, type of line.... If the graphs use relative values, they are preferably modified so that the starting value coincides in the same abscissa, applying the corresponding rule of three.
El gráfico final puede comprender una cuadrícula independiente, que no coincida con la escala de los ejes correspondientes, sino que se defina como proporción de las escalas de los ejes en los gráficos originales.
The final graph may comprise a separate grid, not matching the scale of the corresponding axes, but defined as a proportion of the axis scales in the original graphs.
Claims
REIVINDICACIONES
1- Procedimiento de calibración de gráficos, que ¡guala los ejes de dos o más gráficos (1,2), caracterizado por que se realiza mediante inteligencia artificial y comprende: reconocer las unidades de los ejes de abscisas y ordenadas; comparar y transformar las unidades de ambos ejes; calcular una serie de puntos de cada curva de cada gráfico; redibujar las curvas en un gráfico final (3). 1- Graph calibration procedure, which equalizes the axes of two or more graphs (1,2), characterized in that it is carried out by means of artificial intelligence and includes: recognizing the units of the abscissa and ordinate axes; compare and transform the units of both axes; calculating a series of points from each curve of each graph; redraw the curves in a final graph (3).
2- Procedimiento de calibración de gráficos, según la reivindicación 1, caracterizado por que comprende una etapa previa de localización de gráficos en uno o más documentos mediante inteligencia artificial.
2- Graphics calibration procedure, according to claim 1, characterized in that it comprises a previous stage of locating graphics in one or more documents by means of artificial intelligence.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| PCT/ES2022/070052 WO2023148411A1 (en) | 2022-02-04 | 2022-02-04 | Method for calibrating graphs |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| PCT/ES2022/070052 WO2023148411A1 (en) | 2022-02-04 | 2022-02-04 | Method for calibrating graphs |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| WO2023148411A1 true WO2023148411A1 (en) | 2023-08-10 |
Family
ID=87553176
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| PCT/ES2022/070052 WO2023148411A1 (en) | 2022-02-04 | 2022-02-04 | Method for calibrating graphs |
Country Status (1)
| Country | Link |
|---|---|
| WO (1) | WO2023148411A1 (en) |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| ES2310157A1 (en) * | 2008-06-20 | 2008-12-16 | Universitat De Valencia, Estudi Genera | Compression of colour images, based on non-linear perceptual representations and machine learning |
| ES2329339T3 (en) * | 2000-11-30 | 2009-11-25 | Coppereye Limited | DATABASE. |
| ES2727621T3 (en) * | 2007-02-21 | 2019-10-17 | Palantir Technologies Inc | Provide unique data views based on changes or rules |
-
2022
- 2022-02-04 WO PCT/ES2022/070052 patent/WO2023148411A1/en unknown
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| ES2329339T3 (en) * | 2000-11-30 | 2009-11-25 | Coppereye Limited | DATABASE. |
| ES2727621T3 (en) * | 2007-02-21 | 2019-10-17 | Palantir Technologies Inc | Provide unique data views based on changes or rules |
| ES2310157A1 (en) * | 2008-06-20 | 2008-12-16 | Universitat De Valencia, Estudi Genera | Compression of colour images, based on non-linear perceptual representations and machine learning |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US10663967B2 (en) | Automated driving control device, system including the same, and method thereof | |
| CN111444769B (en) | Laser radar human leg detection method based on multi-scale self-adaptive random forest | |
| CN103400174B (en) | The coded method of a kind of Quick Response Code, coding/decoding method and system | |
| CN104517106B (en) | A kind of list recognition methods and system | |
| US20160351053A1 (en) | Vehicle-to-vehicle beacon message communication system and method | |
| US11074973B2 (en) | Responder signal circuitry for memory arrays finding at least one cell with a predefined value | |
| CN103871407A (en) | Method and apparatus for correcting speech recognition error | |
| CN107742304B (en) | Method and device for determining movement track, mobile robot and storage medium | |
| WO2023148411A1 (en) | Method for calibrating graphs | |
| JP2015158421A (en) | Correction value generation apparatus, fault diagnosis apparatus, correction value generation program, and fault diagnosis program | |
| US20160342158A1 (en) | Method and apparatus for operating a vehicle | |
| AU2020233747B2 (en) | Methods for artificial neural networks | |
| KR20210048450A (en) | Method, apparatus and computer program for generating map including driving route of vehicle | |
| KR20210087496A (en) | Object property detection, neural network training and intelligent driving method, device | |
| HRP20221116T1 (en) | Methods and devices for nucleic acid-based real-time determination of disease states | |
| EP2890041A1 (en) | Space division method, space division device, and space division program | |
| US11334772B2 (en) | Image recognition system, method, and program, and parameter-training system, method, and program | |
| Yang et al. | TAMC: A deep-learning approach to predict motif-centric transcriptional factor binding activity based on ATAC-seq profile | |
| US20210256322A1 (en) | Apparatus and method for classifying attribute of image object | |
| JP7223629B2 (en) | In-vehicle system, external recognition sensor, electronic control unit | |
| CN108846260B (en) | Genetic map construction method and device for genetic segregation population | |
| CN109752730B (en) | Laser positioning method and system based on V-groove detection | |
| CN114252072A (en) | Memory subsystem autonomous vehicle positioning | |
| KR20220009609A (en) | Distance measuring method and device using image tracking for autonomous driving | |
| RU2019132817A (en) | METHOD AND SYSTEM FOR DATA CLASSIFICATION FOR DETECTING CONFIDENTIAL INFORMATION IN TEXT |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| 121 | Ep: the epo has been informed by wipo that ep was designated in this application |
Ref document number: 22924681 Country of ref document: EP Kind code of ref document: A1 |