Method for dynamically adjusting focus point by visual dwell

Abstract

The application relates to a processing method for selecting a focus point by dynamically adjusting a visual stay system, and the method comprises the following steps: (1) defining an interface control element to obtain a focus point; (2) setting the obtained focus point to allow jumping between interface control elements; (3) obtaining a focus point jumping command; (4) loading an automatically obtained initial focus point in the interface of the system, and automatically relocking the focus point when the interface exists a data source refresh; (5) performing focus processing on the obtained focus point; (6) performing data editing processing on the focus point after the focus processing is completed, and determining a specific interaction mode for the current focus point transformation; (7) based on the above operation processing, the focus point transformation is successful, and the dynamic adjustment operation is completed. By adopting the processing method for selecting a focus point by dynamically adjusting a visual stay system, the processing of specific business logic is completed, and the purpose of reducing repeated interaction operations and misoperations can be achieved.

Description

Technical Field

[0001] This invention relates to the field of numerical control interactive control technology, and more particularly to the field of multi-focus control management technology, specifically a method for processing the selected focus of a system by using visual dwell to dynamically adjust the focus. Background Technology

[0002] Currently, the focus control methods commonly used in existing technologies are mainly the following:

[0003] Step 1: Turn on the panel power, start the CNC system, enter the main interactive interface, and the control focus will automatically be concentrated on the default option on the main page;

[0004] Step 2: Switch the interactive interface and adjust the control focus according to the default options set for each interface. For example, when switching to the tool management page, the control focus will also be automatically bound to the tool setting input box number. If you want to switch, you can only manually operate the left and right buttons on the panel to switch.

[0005] Step 3: Each time you manually change the control focus, if you want to enter modified data into the settings text box, you need to reselect the input buffer in the interaction window before you can write to it.

[0006] Step 4: After writing is complete, if you want to switch to the next interactive interface, you can jump directly through the function button. However, the system's control focus will remain on the previous operation screen. You must manually move the control focus to the current interactive interface before you can select the next function.

[0007] In response, when existing technologies adopt the implementation method described in step 2 above, it is necessary to manually control the left and right buttons to shift the control focus. If the interaction is large-scale or infrequent, this type of technical solution can solve the application problem well. However, when the interaction range is small and the operation frequency is high (such as the entry of a large amount of tabular data), this implementation method will bring many risks of misoperation, and the user's interaction control will also be slightly difficult.

[0008] When using the implementation method described in step 3 above, the current UI interaction control technology can only support the selection of a single control by default. This default implementation leads to repetitive interactions, and during data entry, each selection of the modified item and input requires reselection. Given the same probability of error, the increased total number of operations naturally increases the number of incorrect interactions. If the default settings are not used for focus switching, the focus management interface supporting keyboard and mouse input in the UI framework must be called. This means that many automatic focus acquisition mechanisms will be interfered with by the system's default focus transition mechanism. Furthermore, some controls have focus acquisition enabled by default; disabling them if you don't want them to be the focus source is necessary, which is inconvenient for maintenance. Summary of the Invention

[0009] The purpose of this invention is to overcome the shortcomings of the prior art and provide a method for simultaneous multi-focus control, selection, and input, thereby reducing repetitive interactive operations and erroneous operations by using a visual dwell dynamic adjustment system to process the selected focus.

[0010] To achieve the above objectives, the present invention provides a method for selecting the focus using a visual persistence dynamic adjustment system as follows:

[0011] The main feature of this method for dynamically adjusting the selected focus using a visual persistence system is that the method includes the following steps:

[0012] (1) Define a UI control element to receive focus;

[0013] (2) Set the focus of the acquired element to allow navigation between UI control elements;

[0014] (3) Obtain focus jump command, and set the pre-jump for the currently obtained focus;

[0015] (4) Load the automatically acquired initial focus into the system interface and automatically re-lock the focus when the interface is refreshed by a data source;

[0016] (5) Focus the current focus acquired by the system;

[0017] (6) Perform data editing on the focus that has completed the focusing process, and determine the specific interaction method for the current focus change;

[0018] (7) Based on the above operation, the focus change is successful and the dynamic adjustment operation is completed.

[0019] Preferably, step (1) specifically comprises:

[0020] Use additional attributes to mark UI elements, thereby defining a UI control element to receive focus.

[0021] Preferably, step (2) specifically includes:

[0022] (2.1) Define a focus management container control. The focus management container control automatically assigns a sequence number to all focusable elements and focus groups inside the focus management container. This sequence number will be used for program selection when jumping between focus points. When it comes to focus control priority, it can also be used as a sorting index value.

[0023] (2.2) Define the strategy for changing focus using the additional attributes described above;

[0024] (2.3) Each focus group in the currently editable interface control element can define the strategy in step (2.2) above, and the system's focus change strategy is to jump the focus according to the strategy of the lowest focus group that can be changed.

[0025] Preferably, step (2.2) specifically includes:

[0026] The focus change strategy is defined by allowing the focusable elements and focus groups within the focus management container to jump within or across groups from left to right or up to down, and the jump direction can be set.

[0027] Preferably, step (3) specifically includes:

[0028] (3.1) In the focus management container, a delegate command type property is back-bound to the ViewModel of the main interface view, and then called by the ViewModel of the main interface view;

[0029] (3.2) Re-lock the delegate command properties of the currently focused object, or package them into an object and back-bind them to the ViewModel.

[0030] Preferably, step (4) specifically includes:

[0031] (4.1) Find the first focusable element in the management container and reset the focus.

[0032] (4.2) When the system detects that it is refreshing the data source, the focus element is re-locked according to the recorded focus path, so that the focus is on.

[0033] Preferably, after the system completes focus reset and relock focus, the system will recalculate the topology of focus and focus group.

[0034] Preferably, step (5) specifically includes:

[0035] If the additional attribute value of the current focus group is also defined as a focusable element, then the first focusable element of this group will be the focus group itself, thus completing the process of focusing the entire group.

[0036] More preferably, step (6) specifically includes:

[0037] (6.1) Bind the focus path directly in the ViewModel and obtain the data to which the current focus belongs according to the topology structure, and modify it directly; or, attach attributes to the current focusable elements of the system, indicate the name of its key attributes, and directly trigger the delegate to modify the value through the proxy object, send the new value to the focus container, the focus container modifies the key attribute value of the current focus element, and modifies the data in the ViewModel through two-way binding;

[0038] (6.2) Choose to use the left and right buttons on the system panel to control the focus change; or choose to use the camera to capture the user's line of sight to change the focus, thus completing the specific interaction method for changing the focus.

[0039] More preferably, the process of selecting a camera to capture the user's gaze position and perform focus transformation specifically involves: first, locating the iris center and the eye center; after successful iris center location, using a multi-task cascaded neural network to learn the mapping from the relationship between the iris center and the eye center to the gaze point; the transformation process specifically includes:

[0040] A. Based on the MTCNN neural network framework, a multi-task cascaded neural network is designed to locate the iris center and inner and outer canthi. The algorithm consists of two convolutional neural networks.

[0041] B. The first convolutional neural network is used to determine the position of the human eye, obtain the candidate eye region window and its bounding box position, and generate a candidate window for the human eye position;

[0042] C. The human eye region generated by the first convolutional neural network is fed into the second convolutional neural network. The first convolutional neural network is used to determine the iris center and the key point positions of the human eye region, and to mark the iris center.

[0043] D. Based on the above processing, obtain two-dimensional coordinate information of the iris center, inner and outer corners of the eye, and the junction of the iris and upper eyelid on the image. Based on this two-dimensional coordinate information, establish a mapping relationship between it and the gaze target, thereby completing the focus change of capturing the user's gaze position through the camera.

[0044] This invention employs a visual focus-based dynamic adjustment system, utilizing a focus control management mechanism to handle specific business logic through the binding of additional attributes. This provides a method for simultaneous multi-focus control, selection, and input without interference from default focus activation, reducing repetitive interactions and errors. In CNC systems with integrated control panels, hardware requirements are often insufficient, as many settings are modified using left and right keys. The default focus mechanism is difficult to handle, allowing only single focus to be processed at a time (to avoid interference from default focus activation). Therefore, the visual focus capture method provided by this invention solves the problem of misoperation when switching between left and right keys, achieving rapid switching of interactive control focus. Compared to existing technologies, it allows for navigation between interface control elements at different focuses; automatically acquires the initial focus upon interface loading and automatically re-locks the focus when the interface refreshes; prevents focus group boundary overflows; aggregates entire focus groups; and allows simultaneous input and editing of data at the focus point, resulting in a wider range of applications. Attached Figure Description

[0045] Figure 1 This is a flowchart of the method for selecting focus using a visual dwell dynamic adjustment system according to the present invention.

[0046] Figure 2 This is a schematic diagram illustrating the frontal human eye perspective used by the present invention to achieve visual capture of the focal point of the line of sight.

[0047] Figure 3 This is a schematic diagram illustrating the top-down human eye perspective used by the present invention to achieve visual capture of the focal point of the line of sight. Detailed Implementation

[0048] To more clearly describe the technical content of the present invention, the following description is provided in conjunction with specific embodiments.

[0049] Before describing the embodiments of the present invention in detail, it should be noted that, in the following, the terms “comprising,” “including,” or any other variations are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed or inherent to such process, method, article, or apparatus.

[0050] In addition, to make the description of this technical solution clearer and more accurate, the following explanations are provided for the relevant technical terms used in this case:

[0051] 1. Focused element: The currently editable UI control element;

[0052] 2. Focus group: A logical concept consisting of multiple focuses, defining its boundaries and jump rules;

[0053] 3. ViewModel: Responsible for managing the data encapsulation or reading and writing of the business interface View.

[0054] Please see Figure 1 As shown, this method for dynamically adjusting the selected focus using a visual dwell time system includes the following steps:

[0055] (1) Define a UI control element to receive focus;

[0056] (2) Set the focus of the acquired element to allow navigation between UI control elements;

[0057] (3) Obtain focus jump command, and set the pre-jump for the currently obtained focus;

[0058] (4) Load the automatically acquired initial focus into the system interface and automatically re-lock the focus when the interface is refreshed by a data source;

[0059] (5) Focus the current focus acquired by the system;

[0060] (6) Perform data editing on the focus that has completed the focusing process, and determine the specific interaction method for the current focus change;

[0061] (7) Based on the above operation, the focus change is successful and the dynamic adjustment operation is completed.

[0062] In a preferred embodiment of the present invention, step (1) specifically comprises:

[0063] Use additional attributes to mark UI elements, thereby defining a UI control element to receive focus.

[0064] In a specific embodiment of the present invention, the above-mentioned additional attributes are used to mark interface elements, such as adding an isPoint item attribute with a value of type BOOL.

[0065] In a preferred embodiment of the present invention, step (2) specifically includes:

[0066] (2.1) Define a focus management container control. The focus management container control automatically assigns a sequence number to all focusable elements and focus groups inside the focus management container. This sequence number will be used for program selection when jumping between focus points. When it comes to focus control priority, it can also be used as a sorting index value.

[0067] (2.2) Define the strategy for changing focus using the additional attributes described above;

[0068] (2.3) Each focus group in the currently editable interface control elements can define the strategy in step (2.2) above, and the system's focus change strategy is to jump the focus according to the strategy of the focus group at the lowest level of the changeable jump (that is, the focus group below the sequence number does not have the changeable strategy of the current operation).

[0069] In a preferred embodiment of the present invention, step (2.2) specifically includes:

[0070] The focus change strategy is defined by allowing the focusable elements and focus groups within the focus management container to jump within or across groups from left to right or up to down, and the jump direction can be set.

[0071] In a specific embodiment of the present invention, the specific application of step (2) above in jumping between interface control elements that have gained focus includes:

[0072] Define a focus management container control, such as a structure named Point_Manage{}. All focusable elements and focus groups within the focus management container will be automatically assigned a sequence number. In principle, navigation within a group should follow the sequence number, and the selected group's sequence number should remain unchanged after navigation across groups. Whether left / right or up / down navigation triggers navigation within a group or across groups, and the navigation direction, should be configurable. These strategies are defined using attached attributes and can be inherited, so each focus group can redefine these strategies. The focus change strategy follows the strategy of the focus group at the bottom.

[0073] In a preferred embodiment of the present invention, step (3) specifically includes:

[0074] (3.1) In the focus management container, a delegate command type property is back-bound to the ViewModel of the main interface view, and then called by the ViewModel of the main interface view;

[0075] (3.2) Re-lock the delegate command properties of the currently focused object, or package them into an object and back-bind them to the ViewModel.

[0076] In a specific embodiment of the present invention, the specific application of the above-mentioned focus jump command includes:

[0077] Within the focus management container, a delegate command type property (e.g., isChecked = true;) is backbound to the ViewModel of the main view, and then called by the ViewModel of the main view. Similarly, there are delegate command properties for resetting or re-locking focus, or these can be packaged into an object and backbound to the ViewModel.

[0078] In a preferred embodiment of the present invention, step (4) specifically includes:

[0079] (4.1) Find the first focusable element in the management container and reset the focus.

[0080] (4.2) When the system detects that it is refreshing the data source, the focus element is re-locked according to the recorded focus path, so that the focus is on.

[0081] In a preferred embodiment of the present invention, after the system completes focus reset and re-lock focus, the system will recalculate the topology of focus and focus group.

[0082] In practical applications, re-locking and regaining focus includes:

[0083] When resetting focus, look for the first focusable element in the focus management container as the focus.

[0084] When the data source is refreshed, the focus element is re-locked and made to focus by the recorded focus path (the path tracking can be stored in data). (Context locking may result in changes to the context object.)

[0085] Resetting and relocking focus should both require recalculating the topology of focus and focus group.

[0086] In a preferred embodiment of the present invention, step (5) specifically comprises:

[0087] The additional attribute value of the current focus group is also defined as a focusable element, similar to the operation of a pointer pointing to a memory address of a linked list. By default, the pointer initially points to the address of the first element. Therefore, the first focusable element of this group is the focus group itself, thus completing the processing of focusing the entire group. As a result, the focus of the same group can be found directly through the first focus element. Jumping within the group can report status information to the data processing center.

[0088] In a specific embodiment of the present invention, the process of focusing the entire focus group is as follows: the additional attribute value of the focus group is also defined as a focusable element, which can be applied as a new additional attribute value (such as is_point_group). Then the first focusable element of this group is the focus group itself, and the focus group itself also has the is_point additional attribute.

[0089] In a preferred embodiment of the present invention, step (6) specifically includes:

[0090] (6.1) Bind the focus path directly in the ViewModel and obtain the data to which the current focus belongs according to the topology structure, and modify it directly; or, attach attributes to the current focusable elements of the system, indicate the name of its key attributes, and directly trigger the delegate to modify the value through the proxy object, send the new value to the focus container, the focus container modifies the key attribute value of the current focus element, and modifies the data in the ViewModel through two-way binding;

[0091] (6.2) Choose to use the left and right buttons on the system panel to control the focus change; or choose to use the camera to capture the user's line of sight to change the focus, thus completing the specific interaction method for changing the focus.

[0092] In practical applications, when editing data at the focus point, one approach is to directly bind the focus path to the ViewModel, determine the currently focused data based on the topology, and modify it directly. Another approach is to attach attributes to the focusable element, specifying its key attribute name; then, through a delegate object (mentioned in step 3 above), the modification value is triggered, sending the new value to the focus container. The focus container then modifies the key attribute value of the currently focused element, thus modifying the ViewModel's data through two-way data binding.

[0093] Please see Figure 2 As shown, it represents the viewing angle directly facing the human eye. EC represents the center of the eyeball, and OC is the center of the iris. The inner and outer corners of the eye are P1 and P2, respectively, and the junction of the upper eyelid and the iris is U1 and U2. Figure 3This represents the viewpoint from above, with the vector passing through the center of the eyeball and iris denoted by 'd'. Research shows that when the eye gazes in different directions, the iris center rotates around the eyeball center, while the eyeball center remains fixed. Therefore, the direction of gaze can be deduced as the direction of the line connecting the iris center and the eyeball center, i.e., the vector 'd' passing through the iris center, starting from the eyeball center. Since the human eyeball can be considered a sphere, the center of the eyeball can be deduced to be located on the line connecting the two corners of the eye, and equidistant from both corners. Based on this, the position of the eyeball center (EC) can be located. Then, by determining the positions of the iris center (OC) and the two corners of the eye (P1, P2), the direction of gaze can be determined. U1 and U2 are used to assist in locating the iris center and to determine whether squinting occurs during gaze. Figure 3 In this context, vector d can be directly considered as the EC→OC vector, i.e. Figure 2 The vector connecting EC and OC.

[0094] As described above, the camera selects the user's gaze position for focus shifting. Specifically, it involves: first, locating the iris center and the eye center; after successful iris center location, using a multi-task cascaded neural network to learn the mapping from the relationship between the iris center and the eye center to the gaze point. The transformation process specifically includes:

[0095] A. Based on the MTCNN neural network framework, a multi-task cascaded neural network is designed to locate the iris center and inner and outer canthi. The algorithm consists of two convolutional neural networks.

[0096] B. The first convolutional neural network is used to determine the position of the human eye, obtain the candidate eye region window and its bounding box position, and generate a candidate window for the human eye position;

[0097] C. The human eye region generated by the first convolutional neural network is fed into the second convolutional neural network. The first convolutional neural network is used to determine the iris center and the key point positions of the human eye region, and to mark the iris center.

[0098] D. Based on the above processing, obtain two-dimensional coordinate information of the iris center, inner and outer corners of the eye, and the junction of the iris and upper eyelid on the image. Based on this two-dimensional coordinate information, establish a mapping relationship between it and the gaze target, thereby completing the focus change of capturing the user's gaze position through the camera.

[0099] In practical applications, the design of this convolutional neural network structure is as follows:

[0100] The first convolutional neural network consists of four convolutional layers and three pooling layers. The distribution is as follows: the first two convolutional layers are each followed by a pooling layer, the third convolutional layer is directly connected to the fourth convolutional layer, and finally, a pooling layer is connected to the third convolutional layer. Because the input data is a 24×24 image of the human eye region, to reduce computational cost, the size of the first two convolutional and pooling layers is set to 3×3, and the size of the last two layers is set to 2×2. The stride of the convolutional and pooling layers is 1 and 2, respectively.

[0101] The second convolutional neural network consists of two convolutional layers, two pooling layers, and two terminal structures. This network is used to determine the location of key points in the human eye region, thus requiring high accuracy and generalization ability. The first two convolutional and pooling layers are still 3×3 in size, with strides of 1 and 2 respectively. However, since the input image is a 48×48 image of the human eye region, and different experimenters have different eye sizes and shapes, the proportion of the eye region in the input image also varies. To address this, this paper adds terminal structures in the latter half of the network to improve accuracy. The terminal structure consists of three separate 1×1, 3×3, and 5×5 convolutional kernels, followed by a 3×3 pooling layer and then a 1×1 convolutional kernel, arranged side-by-side. This structure allows it to sample data at different scales, extract features from different scales of the image, and finally fuse them to obtain more image features.

[0102] In practical applications, when capturing the user's gaze position and changing the focus using a camera, the gaze tracking algorithm employed in this invention can be broadly divided into two steps: first, iris center and eye center localization; second, learning the mapping from the relationship between the iris center and eye center to the gaze point. Among these, iris center and eye center localization is a crucial part of gaze focus localization. Once the iris center is successfully localized, the mapping method can be used to calculate the gaze point on the screen.

[0103] First, referencing the MTCNN neural network framework, a multi-task cascaded neural network is designed to locate the iris center and the inner and outer canthi. The algorithm consists of two convolutional neural networks, each with the following function:

[0104] ① Determine the location of the human eye and generate candidate windows for that location. This stage uses a convolutional neural network to obtain the candidate eye region windows and their bounding boxes, which is essential for accurately determining the location of the human eye in the input image.

[0105] ② This stage determines the location of key points in the human eye region. The human eye region generated in the previous stage is fed into a second convolutional neural network. This network is more accurate than the first network and can determine the location of key points such as the iris center and the corner of the eye.

[0106] Deep convolutional neural networks (CNNs) are an end-to-end learning approach. For a CNN model, the input data is raw, unprocessed samples, followed by numerous operational layers stacked on top of the input layer. These operational layers can be viewed as a complex function, and the final error function is composed of both data error and the regularization error of the model parameters.

[0107] After marking the center of the human eye iris in the previous step, we can obtain the two-dimensional coordinate information of the iris center, inner and outer canthi, and the junction of the iris and upper eyelid on the image, and establish its mapping relationship with the gaze target. Due to the angular motion of the iris, the mapping relationship of its vector to the screen focus is non-linear, so a second-order polynomial regression model is used to calculate the screen focus coordinates.

[0108] In addition, when this technical solution is used to handle situations where data is modified using the arrow keys (panel controls require the arrow keys to change focus), a focus group is used, which contains only one focusable element. The attached attribute binding command is used to intercept its focus change operation. In this case, the action within the group will inevitably trigger the out-of-bounds command interception, while the cross-group action will be executed normally.

[0109] Any process or method description in the flowchart or otherwise herein can be understood as representing a module, segment, or portion of code comprising one or more executable instructions for implementing a particular logical function or process, and the scope of the preferred embodiments of the invention includes additional implementations in which functions may be performed not in the order shown or discussed, including substantially simultaneously or in reverse order depending on the functions involved, as will be understood by those skilled in the art to which embodiments of the invention pertain.

[0110] It should be understood that various parts of the present invention can be implemented using hardware, software, firmware, or a combination thereof. In the above embodiments, multiple steps or methods can be implemented using software or firmware stored in memory and executed by a suitable instruction execution device.

[0111] Those skilled in the art will understand that all or part of the steps of the methods in the above embodiments can be implemented by a program instructing related hardware. The program can be stored in a computer-readable storage medium, and when executed, the program includes one or a combination of the steps of the method embodiments.

[0112] The storage media mentioned above can be read-only memory, disk, or optical disk, etc.

[0113] In the description of this specification, references to terms such as "an embodiment," "some embodiments," "example," "specific example," or "embodiment," etc., indicate that a specific feature, structure, material, or characteristic described in connection with that embodiment or example is included in at least one embodiment or example of the invention. In this specification, illustrative expressions of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in one or more embodiments or examples.

[0114] Although embodiments of the present invention have been shown and described above, it is understood that the above embodiments are exemplary and should not be construed as limiting the present invention. Those skilled in the art can make changes, modifications, substitutions and variations to the above embodiments within the scope of the present invention.

[0115] This invention employs a visual focus-based dynamic adjustment system, utilizing a focus control management mechanism to handle specific business logic through the binding of additional attributes. This provides a method for simultaneous multi-focus control, selection, and input without interference from default focus activation, reducing repetitive interactions and errors. In CNC systems with integrated control panels, hardware requirements are often insufficient, as many settings are modified using left and right keys. The default focus mechanism is difficult to handle, allowing only single focus to be processed at a time (to avoid interference from default focus activation). Therefore, the visual focus capture method provided by this invention solves the problem of misoperation when switching between left and right keys, achieving rapid switching of interactive control focus. Compared to existing technologies, it allows for navigation between interface control elements at different focuses; automatically acquires the initial focus upon interface loading and automatically re-locks the focus when the interface refreshes; prevents focus group boundary overflows; aggregates entire focus groups; and allows simultaneous input and editing of data at the focus point, resulting in a wider range of applications.

[0116] In this specification, the invention has been described with reference to specific embodiments thereof. However, it will be apparent that various modifications and variations can be made without departing from the spirit and scope of the invention. Therefore, the specification and drawings should be considered illustrative rather than restrictive.

Claims

1. A method for selecting a focus using a visual persistence dynamic adjustment system, characterized in that, The method includes the following steps: (1) Define a UI control element to gain focus; (2) Set the focus of the acquired element to allow navigation between UI control elements; (3) Obtain focus jump command, and set the pre-jump for the currently obtained focus; (4) Load the automatically acquired initial focus into the system interface, and automatically re-lock the focus when the interface is refreshed by a data source; (5) Focus the current focus acquired by the system; (6) Perform data editing on the focus that has completed the focusing process, and determine the specific interaction method for the current focus change; (7) Based on the above steps, the focus change is successful, and the dynamic adjustment operation is completed; The specific steps (1) are as follows: Use additional attributes that bind focus information to mark UI elements. This mark can be used to determine whether the current control element has a focus lock and whether it can be used to select the jump focus, thereby defining a UI control element to get focus. Step (2) specifically includes: (2.1) Define a focus management container control. The focus management container control automatically assigns a sequence number to all focusable elements and focus groups inside the focus management container. This sequence number will be used for program selection when jumping between focus points. When focus control priority is involved, it can also be used as a sorting index value. (2.2) Define the strategy for changing focus using the additional attributes described above; (2.3) Each focus group in the currently editable interface control element can define the strategy in step (2.2) above, and the system's focus change strategy is to jump the focus according to the strategy of the lowest focus group that can be changed; After the system completes focus reset and relock, the system will recalculate the topology of the focus and focus group. Step (6) specifically includes: (6.1) Bind the focus path directly in the main interface view ViewModel, and obtain the data to which the current focus belongs according to the topology structure, and modify it directly; or, attach attributes to the current focusable elements of the system, indicate the name of its key attributes, and directly trigger the delegate to modify the value through the proxy object, send the new value to the focus container, the focus container modifies the key attribute value of the current focus element, and modifies the data in the ViewModel through two-way binding; (6.2) Select to use the left and right buttons on the system panel to control the focus change; or select to capture the user's line of sight with the camera to change the focus, thereby completing the specific interaction method for changing the focus; The aforementioned selection of the camera to capture the user's gaze position and perform focus transformation specifically involves: firstly, locating the iris center and the eye center; after the iris center is successfully located, using a multi-task cascaded neural network to learn the mapping from the relationship between the iris center and the eye center to the gaze point; the transformation process specifically includes: A. Based on the MTCNN neural network framework, design a multi-task cascaded neural network to locate the iris center and inner and outer canthi of the eye. The algorithm consists of two convolutional neural networks. B. The first convolutional neural network is used to determine the position of the human eye, obtain the candidate eye region window and its bounding box position, and generate a candidate window for the human eye position; C. The human eye region generated by the first convolutional neural network is fed into the second convolutional neural network. The first convolutional neural network is used to determine the iris center and the key point positions of the human eye region, and to mark the iris center. D. Based on the above processing, obtain two-dimensional coordinate information of the iris center, inner and outer corners of the eye, and the junction of the iris and upper eyelid on the image. Based on this two-dimensional coordinate information, establish a mapping relationship between it and the gaze target, thereby completing the focus change of capturing the user's gaze position through the camera.

2. The method for selecting focus using a visual persistence dynamic adjustment system according to claim 1, characterized in that, Step (2.2) specifically includes: The focus change strategy is defined by allowing the focusable elements and focus groups within the focus management container to jump within or across groups from left to right or up to down, and the jump direction can be set.

3. The method for selecting focus using a visual persistence dynamic adjustment system according to claim 2, characterized in that, Step (3) specifically includes: (3.1) In the focus management container, a delegate command type property is back-bound to the ViewModel of the main interface view, and called by the ViewModel of the main interface view; (3.2) Re-lock the delegate command properties of the currently focused object, or package them into an object and back-bind them to the ViewModel.

4. The method for selecting focus using a visual persistence dynamic adjustment system according to claim 3, characterized in that, Step (4) specifically includes: (4.1) Find the first focusable element in the management container and reset the focus. (4.2) When the system detects that it is refreshing the data source, the focus element is re-locked according to the recorded focus path, so that the focus is on.

5. The method for selecting focus using a visual persistence dynamic adjustment system according to claim 4, characterized in that, The specific steps (5) are as follows: If the additional attribute value of the current focus group is also defined as the focusable element, then the first focusable element of this group is the focus group itself, thus completing the process of focusing the entire group.

Images