TWM592541U - Image recognition system - Google Patents

Abstract

An image recognition system includes a camera device, an acquisition module, an image processing module, a face database, a financial product database, and an arithmetic module. The capturing module captures sub-images including one of the user's facial features from the images obtained by the camera device and calculates measurable information accordingly. The measurable information includes a straight line or area composed of a plurality of feature points. The face database stores attributes and evaluation values of measurable information. The financial product database stores a plurality of financial product information and age ranges according to risk ratio ranges. The calculation module estimates the user's age based on the measurable information, obtains the evaluation value from the face database according to the attribute corresponding to the measurable information and calculates the risk value based on it, and obtains a value from the financial product database based on the risk value and age Financial product.

Description

Image recognition system

This creation is about an image recognition system, especially an image recognition system that obtains facial features from images.

For example, financial institutions such as banks and post offices set up multiple surveillance cameras in their business halls to monitor people traveling in and out of the business halls.

However, the surveillance images taken by these surveillance cameras are usually used to review afterwards special events that occurred at a specific point in time (such as masked robbery, driver theft, etc.). In other words, the monitoring system set up by financial institutions lacks a mechanism for instant processing and reporting based on current images. Although the monitoring system has been running for a long time, it is only used as a safety device that may deter people who are interested. In addition, the overhead surveillance screen from a height must include all kinds of people and objects, such as waiting people, counter people, counter clerks, lobby floors, waiting seats, ATMs, folding machines, etc.; even if Arranging a dedicated person to watch the surveillance image and quickly lock the face of a specific person from a complex picture is not easy, let alone that humans are prone to miss important information in the picture due to fatigue or distraction.

In addition, for high-net-worth customers who frequently come and go, financial institutions provide exclusive financial services to customers at this customer level. Through customized commodity investment portfolios, high-net-worth customers can increase their wealth. On the other hand, financial institutions usually assist clients who are generally willing to carry out wealth management through mid-term and long-term asset planning through medium- and long-term asset planning. By understanding the client’s current financial flow and future capital use planning, they conduct risk assessments to propose a complete financial plan. Product recommendation plan.

However, for customers who simply come to the business hall to handle general financial business, the clerks or financial specialists of the financial institution cannot grasp whether the customer is a potential customer with financial needs based on the facial features of the customer. Therefore, the clerks are promoting financial products Time is often formalized, unable to provide financial products that may meet the needs of this type of customer, and can only blindly promote projects that the customer is not interested in, wasting both parties' time and energy in vain.

In view of this, this creation proposes an image recognition system that can obtain face images from the images captured by surveillance cameras and further analyze the measurable information in the facial features, thus improving the sales success rate of financial products and reducing customer factors. Probability of being in a bad mood by being recommended as an unsuitable product.

An image recognition system according to an embodiment of the present invention includes a camera device, an acquisition module, an image processing module, a face database, a financial product database, and an arithmetic module. The imaging device can obtain images. The capture module selects a block from the image and obtains a face image of a user in the block. The image processing module is electrically connected to the capture module. The image processing module captures the sub-image based on the face image and calculates measurable information based on the sub-image. The sub-image includes one of the user's facial features and the surrounding skin. The measurable information includes a straight line or an area composed of a plurality of feature points. The face database includes a plurality of first table spaces, and each first table space corresponds to one of the five sense types and includes a plurality of first fields. Each first field stores attributes and evaluation values of measurable information. The attributes include the length of the line, the shape and color of the area. The financial product database includes a plurality of second table spaces. These second table spaces respectively correspond to a plurality of risk ratio intervals and include a plurality of second fields. Each second field stores a financial product and one-year interval. The computing module communicates with the image processing module, the face database and the financial product database. The computing module estimates one year of the user based on the measurable information, obtains the evaluation value from the face database according to the attribute corresponding to the measurable information, calculates the risk value based on the evaluation value, and based on the risk ratio interval corresponding to the risk value One selects a second table space from the financial product database, selects and outputs a second field from the selected second table space according to the age interval corresponding to the age.

With the above structure, the image recognition system disclosed in this case can quickly lock the face image in a specific area from the complex image taken by the surveillance camera, and analyze the age and facial features of the face image through face recognition Measurable information, and then convert multiple measurable information into an evaluation score. Therefore, when a customer visits a specific counter in the business hall of a financial institution, the image recognition system disclosed in this case can further recommend suitable financial products based on the range of the evaluation score.

The above description of the content of the disclosure and the following description of the embodiments are used to demonstrate and explain the spirit and principle of the creation, and provide a further explanation of the scope of the patent application of the creation.

The following describes in detail the detailed features and advantages of the creation in the embodiments. The content is sufficient for any person skilled in the relevant art to understand and implement the technical content of the creation, and according to the content disclosed in this specification, the scope of patent application and the drawings Anyone who is familiar with related skills can easily understand the purpose and advantages of this creation. The following examples further illustrate the views of this creation in detail, but do not limit the scope of this creation with any views.

Please refer to FIG. 1, which is a block diagram of an image recognition system 10 according to an embodiment of the present invention. The image recognition system 10 described above is suitable for installation in the business hall of financial institutions. The image recognition system 10 includes: a camera device 1, an acquisition module 2, an image processing module 3, a face database 4, a financial product database 5, and an arithmetic module 6.

The imaging device 1 can acquire images. Please refer to FIG. 2, which is a schematic top view of the camera device 1 installed in the business hall of a financial institution. As shown in FIG. 2, the location of the camera device 1 can capture the surrounding environment of a specific counter in the business hall of a financial institution. Please refer to FIG. 3, which illustrates an image captured by the camera device 1 provided in FIG. 2. The image captured by the camera device 1 includes users around the counter.

Please refer to Figure 1. The capture module 2 is electrically connected to the camera device 1. The capturing module 2 can perform color space conversion, color filtering and edge detection according to a block to obtain a face image. In detail, please refer to Figure 3 together. The system administrator can select a region of interest (ROI) from the monitoring screen as the aforementioned block in advance, thereby saving the cost of capturing the image outside the region of interest ROI by the capturing module 2. In practice, the ROI of the region of interest in the screen can be preset according to the perspective of the counter service personnel, as shown in FIG. 3.

After setting the block based on the region of interest ROI, the capture module 2 can convert the RGB color space of the block into one of the HSV color space, CIE 1931 color space, YIQ color space, and YCbCr color space, thereby reducing light For the impact of subsequent face detection. After the color space conversion is completed, the capturing module 2 separates the skin color from the background color. In detail, the distribution of skin color in the converted color space can be defined by a specified system of equations. In addition, because the color of the background objects in the business hall of financial institutions is fixed, the background color can be defined by another specified equation group. The capture module 2 uses, for example, Otsu Thresholding or Balanced Histogram Thresholding (BHT) to find all pixels in the block that meet the skin color condition. At the same time, the capture module 2 can And filter out pixels with background color to speed up the overall processing speed.

In an embodiment, for the binarized image, the capture module 2 may use a dilation operator and an erosion operator on the morphology to eliminate noise outside the face area. Specifically, the capturing module 2 can perform a closing operation of expanding first and then corroding (Closing operation) and then an opening operation of first corroding and then expanding (Open operation), thereby highlighting the contours around the facial features in the image.

For the image block after noise removal, the capturing module 2 selects the face image in the block according to the geometric model of the face. The geometric model can be defined according to the interval of the ratio of the height and width of a general face, for example: 0.79≤face height/face width≤2.95.

In another embodiment, the capturing module 2 uses, for example, a Canny filter, a Sobel filter or a Prewitt operator, thereby performing edge detection of a face image, and further outputting a person Face image.

In yet another embodiment, the capture module 2 further divides the block image based on the YUV color coding method into a plurality of overlapping small areas, and calculates each small size for the brightness value (that is, the Y element) in these areas The edge intensity values of the area in the horizontal direction, the vertical direction and the two diagonal directions are compared with each other according to the calculated intensity values and a threshold value to obtain the face edge, and further output the face image.

In still another embodiment, the capturing module 2 performs comparison based on the image samples of the eigenface trained in advance, and then outputs a face image. In practice, financial institutions can use image files with different faces retained when many customers open an account to perform Principal Component Analysis (Principal Component Analysis, PCA) to obtain a set of vectors of characteristic faces as the aforementioned image samples.

In practice, the extraction module 2 includes, for example, a digital signal processor, an application specific integrated circuit (ASIC), a digital logic circuit, and a field programmable gate array. FPGA) or other hardware components that can perform the above-mentioned facial image capture function.

The image processing module 3 is electrically connected to the capturing module 2. The image processing module 3 obtains a plurality of feature points according to the face image, and captures sub-images according to the distribution of these feature points, and then calculates measurable information based on the sub-images.

In one embodiment, the image processing module 3 uses a connected-component labeling algorithm to find pixels connected in the image so as to mark the coordinates of the facial features in the face image as feature points.

In another embodiment, the image processing module 3 uses an active shape model (Active Shape Model) or an active appearance model (Active Appearance Models, AAM) to calculate the coordinates of facial feature points in the face image at a time.

The sub-image captured by the image processing module 3 according to the feature points includes one of the user's facial features and the surrounding skin. In one embodiment, the sub-image is a first sub-image, and the first sub-image includes the user's eyes and surrounding skin. In another embodiment, the image processing module 3 further captures the second sub-image, the third sub-image, the fourth sub-image, and the fifth sub-image respectively according to the face image. The second sub-image includes the user's eyebrows and surrounding skin, the third sub-image includes the user's nose and surrounding skin, the fourth sub-image includes the user's mouth and surrounding skin, and the fifth sub-image includes the user's ear.

After obtaining the sub-image, the image processing module 3 calculates measurable information according to the sub-image. The measurable information includes a straight line or a region and a region composed of a plurality of feature points. Measurable information includes eye shape, binocular distance, eyebrow length, eyebrow distance, nose width, nose bridge length, upper lip thickness, lower lip thickness, ear shape, ear color, and wrinkle length, wrinkle depth and skin around the facial features Waiting time.

The face database 4 includes a plurality of first table spaces, each of which corresponds to one of the five sense types and includes a plurality of first fields. Specifically, each table space corresponds to a type of the above measurable information. For example, table space A is associated with eye shape, and table space B is associated with binocular distance. Each first field stores the attributes, evaluation values, and a weight value of the measurable information. The attributes include the length of the line, the shape and color of the area. According to the previous example, the attributes stored in each field of table space A correspond to different eye shapes, such as: apricot eyes, danfeng eyes, squint eyes, slender eyes, round eyes, vertical eyes, triangular eyes, etc., each eye shape corresponds to To an evaluation value. The attributes stored in each field of the table space B correspond to different binocular distances, such as 1.1~1.3 cm, 1.3~1.5 cm, 1.5~1.7 cm, etc.

The evaluation values and weight values of the multiple first fields of the face database 4 are obtained by running a deep learning algorithm through another computing module in advance. In detail, financial institutions can use big data such as the original customer image database, investor risk attribute analysis questionnaires filled out by customers, and financial product projects invested by customers as original training sets, through recursive neural networks (Recursice Neural Network, RNN) or Long Short Term Memory (LSTM) obtains the specific facial features corresponding to various financial product types, and then calculates the evaluation value and weight value stored in each field. The evaluation value is related to the proportion of risk that investors are willing to bear, and the weight value represents the proportion of each type that should be considered when considering various measurable information.

The financial product database 5 includes a plurality of second table spaces. These second table spaces respectively correspond to a plurality of risk ratio intervals and include a plurality of second fields, and each second field stores a financial product and a one-year interval. For example, the risk ratio range corresponding to table space C is 0~6, and its risk attributes are typically classified as "conservative", which means that investors are basically reluctant to bear any investment risks. The suitable financial products are biased in compensation from interest income. Product; the risk ratio range corresponding to table space D is 7~13, and its risk attributes are typically classified as "very cautious", which means that investors can basically accept slight losses in exchange for slight potential investment returns, suitable financial products Projects rated as low risk by financial institutions. It should be noted that the boundary value of the above risk ratio interval is obtained by running a deep learning algorithm through another computing module in advance.

The computing module 6 is in communication with the image processing module 3, the face database 4 and the financial product database 5. The computing module 6 estimates the age of the user based on measurable information. In detail, the calculation module 6 mainly estimates the age based on the measurable information such as the wrinkle length, wrinkle depth and skin shift point around the facial features, especially the headlift above the eyebrows, the decree lines on both sides of the nose and Crow's feet around the eyes. The image processing module 3 uses the Canny edge detector technology to detect the texture of wrinkles and set an appropriate threshold value for age judgment. On the other hand, since the skin of the human face will show different shift characteristics as it ages, it is possible to segment the skin color of the skin block of the human face image as an auxiliary method for recognizing age. Specifically, by appropriately selecting the critical values of the parameters in the Y, C _b and _Cr color spaces, and detecting the face image output by the capture module 2, age spots can be detected. For example, if a specific block of a face image meets the following formula: (24≤Y≤124) and (113≤C _b ≤140) and (115≤C _r ≤143), it can be determined that the specified block has Age spots. The computing module 6 can further estimate the age of the user based on the color and number of age spots and the number of head-up lines, decree lines and crow's feet lines.

The computing module 6 obtains one or several evaluation values from the face database 4 according to the attributes corresponding to the measurable information of the five senses, calculates the risk value according to the evaluation value, and one of the risk ratio ranges corresponding to the risk value is from the financial A second table space is selected in the product database 5, and a second field is selected from the selected second table space according to the age interval corresponding to the age and output. In practice, the computing module 6 obtains five evaluations from the surface database 4 based on the attributes corresponding to the measurable information of the first sub-image, the second sub-image, the third sub-image, the fourth sub-image, and the fifth sub-image. Value and five weight values, and the risk value is calculated based on the five evaluation values and the five weight values. In detail, the computing module 6 first obtains the risk assessment values according to the characteristics of the user's facial features (for example, the corresponding risk assessment values of the eyes, eyebrows, noses, and ears are A ₁ , A ₂ , A ₃ , A ₄ and A ₅ ), and then according to the corresponding weight values of the facial features (for example, A ₁ corresponds to W ₁ , A ₂ corresponds to W ₂ …, A ₅ corresponds to W ₅ ), and then these risk assessment values are summed up according to their corresponding weight values Obtain a comprehensive risk assessment value of the user (for example, A ₁ *W ₁ +A ₂ *W ₂ +A ₃ *W ₃ +A ₄ *W ₄ +A ₅ *W ₅ ). The computing module 6 obtains a list of corresponding financial products in the financial product database 5 based on the comprehensive risk assessment value, and then obtains a recommended financial product according to the user age estimated by the computing module.

Based on the above, the image recognition system disclosed in this creation can quickly lock the face image in a specific area from the complex image captured by the surveillance camera, and analyze the age and facial features of the face image through face recognition The measurable information of, and then convert the multiple measurable information into an evaluation score. Therefore, when a customer visits a specific counter in the business hall of a financial institution, the image recognition system disclosed in this case can further recommend suitable financial products based on the range of the evaluation score.

Although this creation is disclosed above with the foregoing embodiments, it is not intended to limit this creation. Without departing from the spirit and scope of this creation, all changes and retouching are within the scope of patent protection of this creation. For the protection scope defined by this creation, please refer to the attached patent application scope.

10: Image recognition system 1: camera device 2: Capture module 3: Image processing module 4: Face database 5: Financial product database 6: arithmetic module ROI: area of interest

FIG. 1 is a schematic diagram of an image recognition system according to an embodiment of the present invention. FIG. 2 is a schematic top view of a camera device installed in a business hall of a financial institution according to an embodiment of the present invention. FIG. 3 is a schematic diagram of images captured by a camera device according to an embodiment of the present invention.

10: Image recognition system

1: camera device

2: Capture module

3: Image processing module

4: Face database

5: Financial product database

6: arithmetic module

Claims (9)

An image recognition system includes: a camera device for obtaining an image; an acquisition module for communicating with the camera device, the acquisition device for selecting a block from the image and obtaining a use in the block One of the face images; an image processing module that communicates with the capture module, the image processing module captures a sub-image based on the face image and calculates measurable information based on the sub-image, wherein the sub-image The image includes one of the user’s facial features and the surrounding skin, and the measurable information includes a straight line or an area composed of a plurality of feature points; a face database includes a plurality of first table spaces, each of which A table space corresponds to one of the five sense types and includes a plurality of first fields, each of which stores an attribute of the measurable information and an evaluation value, where the attribute includes a length of the straight line , A shape and a color of the area; a financial product database, including a plurality of second table spaces, the second table spaces correspond to a plurality of risk ratio intervals and include a plurality of second fields, each These second fields store a financial product and a one-year interval; and an arithmetic module that communicates with the image processing module, the face database, and the financial product database, and the arithmetic module estimates based on the measurable information An age of the user obtains the evaluation value from the face database according to the attribute corresponding to the measurable information, calculates a risk value based on the evaluation value, and based on the risk ratio intervals corresponding to the risk value One selects one of the second table spaces from the financial product database, selects and outputs a second field from the selected second table space according to the age interval corresponding to the age. The image recognition system according to claim 1, wherein the capturing module is further used for color space conversion, color filtering and edge detection according to the block to obtain the face image. The image recognition system according to claim 1, wherein the sub-image is a first sub-image, and the first sub-image includes the user's eye and surrounding skin; the image processing module is further based on the face image Capture a second sub-image, a third sub-image, a fourth sub-image and a fifth sub-image, the second sub-image includes the user's eyebrows and surrounding skin, the third sub-image includes the use The nose and surrounding skin of the user, the fourth sub-image includes the user's mouth and surrounding skin, the fifth sub-image includes the user's ear; the measurable information includes the shape of the eyes, the distance between the eyes, the eyebrows Length, eyebrow spacing, nose width, nose bridge length, upper lip thickness, lower lip thickness, ear shape, ear color, wrinkle length, wrinkle depth, skin shift point; each of the first fields of the face database further stores the A weight value of measurable information; and the arithmetic module according to the measurable value of each of the first sub-image, the second sub-image, the third sub-image, the fourth sub-image and the fifth sub-image The attribute corresponding to the information obtains five evaluation values and five weight values from the face database, and calculates the risk value based on the five evaluation values and the five weight values; wherein, the first fields of the face database The evaluation values and the weight values, as well as the boundary values of the risk ratio intervals of the financial product database, are obtained by running a deep learning algorithm through another computing module in advance. The image recognition system according to claim 1, wherein the capture module is further used to convert one of the color spaces of the block into one of the HSV color space, the CIE 1931 color space, the YIQ color space, and the YCbCr color space. The image recognition system according to claim 1, wherein the capturing module is further used to select the face image in the block based on a face geometric model. The image recognition system according to claim 1, wherein the capture module is further used for edge detection based on the Kanney operator, Sauber operator or Prewitt operator to output the face image. The image recognition system according to claim 1, wherein the capture module compares a feature face sample to output the face image. The image recognition system according to claim 1, wherein the image processing module calculates the coordinates of the facial feature points in the face image according to an active shape model or an active appearance model. The image recognition system according to claim 3, wherein the deep learning algorithm is a long-short-term memory neural network or a recursive neural network.

Images