Man-machine verification method and system based on physical lever balance principle

Abstract

The invention discloses a man-machine verification method and system based on a physical lever balance principle. The invention simulates a physical lever, and provides a visual and convenient verification method. Specifically, key physical parameters (such as net torque d) are hidden, and logic association between the key physical parameters and interface visual elements is cut off, so that automatic cracking based on machine vision or an image recognition technology is eradicated fundamentally. And for general users, the verification method provided by the invention accords with experience intuition, does not need learning, and is high in verification speed. Furthermore, a dual verification mechanism of'balance judgment + behavior characteristic judgment 'is introduced, and scripts and robot behaviors simulating human operation are effectively recognized and intercepted.

Description

Technical Field

[0001] This invention relates to the field of network security technology, and more specifically, to a human-computer verification method and system based on the principle of physical lever balance. Background Technology

[0002] With the rapid development of internet technology, CAPTCHA (Human-to-Machine Verification) technology has become a key security mechanism for distinguishing human users from automated programs, and is widely used in scenarios such as user registration, login, password modification, and transaction confirmation. Currently, the mainstream CAPTCHA technologies mainly include the following categories:

[0003] Image character recognition CAPTCHAs, image dot selection CAPTCHAs, and image puzzle CAPTCHAs have the following significant problems with existing CAPTCHA technologies:

[0004] Machine vision attacks are highly effective. With the development of deep learning technologies (such as convolutional neural networks and optical character recognition), the machine recognition accuracy of traditional image-distorted text CAPTCHAs and image point selection CAPTCHAs has continued to improve. Attackers can use models to achieve batch cracking. There is a deterministic mapping relationship between visual information and answers. The correct answer to existing CAPTCHAs can be directly or indirectly derived from the presented visual information. For example, the answer to a character CAPTCHA is the character in the image; the answer to a jigsaw puzzle CAPTCHA is the coordinate position of the target image. This deterministic mapping relationship allows machines to directly deduce the correct answer. User experience and security are difficult to balance. To improve resistance to cracking, CAPTCHAs often increase difficulty by adding visual interference (such as noise and distortion), which increases the recognition burden on human users and degrades the user experience. Summary of the Invention

[0005] This invention overcomes the shortcomings of existing technologies that use image verification, such as ease of attack and poor user experience, and provides a human-computer verification method and system based on the principle of physical lever balance, in order to solve the problems existing in the prior art.

[0006] To address the aforementioned technical problems, this invention provides a human-machine verification method based on the principle of physical lever balance:

[0007] A human-computer verification method based on the principle of physical lever balance includes the following steps:

[0008] The verification server randomly generates a set of lever physical parameters and sends the lever physical parameters to the verification operation terminal;

[0009] After receiving the physical parameters of the lever, the verification operation terminal renders and generates a lever verification code animation.

[0010] Users interact with the lever verification code animation on the verification terminal;

[0011] If the customer's actions during the verification code animation meet the set requirements, the verification will pass.

[0012] If the set requirements are not met, the verification will be deemed a failure.

[0013] A further technical solution is that the physical parameters of the lever include one of the following:

[0014] Option 1:

[0015] The physical parameters of the lever include the mass of the item on the left, the mass of the item on the right, the mass of the sliding rider, the length of the left lever arm, the length of the right lever arm, and the net torque.

[0016] Option 2:

[0017] The physical parameters of the lever include the force applied on the left, the force applied on the right, the length of the left lever arm, the length of the right lever arm, and the net torque.

[0018] A further technical solution is that, in Solution 1, the generated animation includes a balance scale animation;

[0019] In Option 2, the generated animations include crowbar animation, seesaw animation, and pulley animation.

[0020] A further technical solution is that the customer's operation of the leverage verification code animation meets the set requirements, specifically as follows:

[0021] In Scheme 1, at least one of the physical parameters used in the lever verification code animation rendered by the verification operation terminal is an unbalanced parameter. The user adjusts the unbalanced parameter to make the balance lever reach a balanced state.

[0022] In Scheme 2, at least one of the physical parameters used in the lever verification code animation rendered by the verification operation terminal is a non-set parameter. The user adjusts the non-set parameter to make the crowbar, seesaw, and pulley lever meet the set requirements.

[0023] A further technical solution is that, in the lever physical parameters of the first solution, the net torque is a non-equilibrium parameter;

[0024] The net torque is expressed and adjusted using the position of the balance rider.

[0025] A further technical solution involves the user interacting with the lever verification code animation on the verification terminal, specifically including the following steps;

[0026] Users click and / or swipe on elements in the interface;

[0027] Based on the user's actions, the system generates physical parameters for lever verification and further renders the animation based on these parameters to generate user feedback.

[0028] A further technical solution is that the failure to meet the set requirements specifically means that the balance adjustment cannot be completed and / or the conditions for real-person verification are not met:

[0029] The conditions for real-person verification include: the number of operations exceeds a set number;

[0030] The operation time exceeded the set time;

[0031] The operation trajectory does not conform to the requirements of a real person's operation trajectory.

[0032] A further technical solution is that the determination of failure to meet the set requirements includes the following steps:

[0033] Set the criteria for real-person verification and assign a weight Wi to those criteria.

[0034] Set the conditions and values ​​for real-person verification;

[0035] The value that meets the verification conditions is obtained based on the user's usage characteristics;

[0036] Sum the values ​​of different conditions after multiplying them by their corresponding weights;

[0037] The summed judgment value is compared with the set judgment threshold to determine whether the set requirements are met.

[0038] Another aspect of the present invention provides a human-machine verification system based on the principle of physical lever balance, including a verification server and a verification operation terminal;

[0039] The verification server is used to generate a set of lever physical parameters and send the lever physical parameters to the verification operation terminal;

[0040] The verification terminal is used to render and generate a leverage verification code animation, allowing customers to interact with the animation and determine whether the customer has passed the verification.

[0041] Compared with existing technologies, this invention has at least the following beneficial effects: This invention simulates a physical lever, providing an intuitive and secure verification method. Specifically, this application fundamentally eliminates automated cracking based on machine vision or image recognition technology by hiding key physical parameters (such as net torque d) and severing their logical association with interface visual elements. For general users, the verification method of this application aligns with experience and intuition, requires no learning, and is fast. Furthermore, the introduction of a dual verification mechanism of "balance judgment + behavioral feature judgment" effectively identifies and intercepts scripts and robot behaviors that simulate human operations. Enhanced anti-forgery capability: Assigning the highest weight to trajectory dynamics significantly increases the difficulty for machines to simulate human nonlinear motion patterns. Clear and efficient structure: The three indicators characterize the operation from the perspectives of "quantity," "time," and "quality," respectively, and make a final decision through weighted fusion, giving the system both robustness and interpretability. Attached Figure Description

[0042] Figure 1 This is a schematic diagram of the rendered interface;

[0043] Figure 2 A timing diagram for the human-machine verification method of the lever balance principle;

[0044] Figure 3 A schematic diagram of the human-machine verification method for the lever balance principle;

[0045] Figure 4 This is a schematic diagram of a human-machine verification method system for the lever balance principle. Detailed Implementation

[0046] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the invention.

[0047] Example 1

[0048] A human-computer verification method based on the physical lever balance principle, see [link to relevant documentation]. Figure 2 , Figure 3 This includes the following steps:

[0049] S1: The verification server randomly generates a set of lever physical parameters and sends the lever physical parameters to the verification operation terminal;

[0050] Understandably, the physical parameters of a lever can change depending on the type of lever. For example, in a balance lever scenario, the physical parameters of the lever include the mass of the item on the left, the mass of the item on the right, the mass of the sliding rider, the length of the left lever arm, the length of the right lever arm, and the net torque.

[0051] In typical scenarios involving crowbars, seesaws, and pulleys, the physical parameters of a lever include the force applied to the left, the force applied to the right, the length of the left lever arm, the length of the right lever arm, and the net torque.

[0052] Here, we take the scenario of a balance scale lever as an example. See Figure 1 :

[0053] The generated lever physical parameters are a set of physical parameters for a balance scale, which satisfy the physical equations for lever equilibrium:

[0054] d r = (W l *L l - W r *L r ) / W;

[0055] Among them, W l The mass of the item on the left;

[0056] W r The mass of the item on the right;

[0057] W represents the mass of the sliding rider;

[0058] L l This is the length of the left lever arm;

[0059] L r This is the length of the left lever arm;

[0060] d r The net torque is represented by the position of the sliding rider in a balance.

[0061] S2: After receiving the lever's physical parameters, the verification terminal renders and generates a lever verification code animation. (See below) Figure 1 ;

[0062] Understandably, in the balance scale scenario, the user's adjustment is to move from an unbalanced state to a set balanced state;

[0063] At this point, although the generated parameters are balanced parameters, at least one of the lever physics parameters used for rendering is an unbalanced lever physics parameter.

[0064] Specifically, the modification could be that after the verification server generates the balance parameters, it replaces the net torque in the balance parameters with unbalanced parameters and sends them to the verification operation terminal for rendering; or the verification operation terminal receives the balance parameters and adjusts the net torque to unbalanced parameters for rendering.

[0065] In one example, the verification server generates the following balance parameters:

[0066] W l= 6 (Mass of the item on the left)

[0067] W r = 4 (Mass of the item on the right)

[0068] W = 3 (slipper mass)

[0069] L l = 5 (Length of the left lever arm)

[0070] L r = 7 (length of the right lever arm)

[0071] d r ≈ 0.667 (net torque).

[0072] According to the lever balance formula W l * L l = W r * L r + W * d, calculate the unique solution d that balances the lever. r = (W l *L l - W r *L r ) / W = (6*5 - 4*7) / 3 = (30-28) / 3 ≈ 0.667.

[0073] For example, the verification server will d r The unbalanced parameters are replaced and sent to the verification operation terminal. The verification operation terminal renders the rider position of the balance according to the unbalanced parameters. At this time, the generated balance is in an unbalanced state.

[0074] S3: The user interacts with the lever verification code animation on the verification operation terminal;

[0075] The user's interaction with the lever verification code animation on the verification terminal specifically includes the following steps:

[0076] S31: The user clicks and / or swipes on elements in the interface;

[0077] S32: Generate lever verification physical parameters based on user actions, and further render the animation based on the lever verification physical parameters to generate user action feedback.

[0078] Understandably, in the balance lever scenario, the rider can be moved by sliding; in the crowbar scenario, the force applied by the crowbar can be increased by sliding; and in the balance scale scenario, the rider can be increased by clicking on one end of the tray. The interaction method is not limited here.

[0079] exist Figure 1 In the middle, specific prompts are generated, and dragging "Wukong" or the "bottom slider" balances the scale, allowing customers to understand the specific operation methods expressed on the interface.

[0080] S4: If the customer's actions during the lever verification code animation meet the set requirements, the verification will pass.

[0081] If the set requirements are not met, the verification will be deemed a failure.

[0082] Taking the adjustment from a non-equilibrium state to a set equilibrium state as an example, when d r = (W l *L l - W r *L r When W = 0.5, equilibrium is reached;

[0083] In this embodiment, d r = (W l *L l - W r *L r ) / W = (6*5 - 4*7) / 3 = (30-28) / 3 ≈0.667;

[0084] For example, considering the precision of user operations, the system adopts a dual fault-tolerance mechanism: absolute error E abs Less than 0.5, or relative error E rel If the percentage is less than 70%, the balance is considered successful. An "OR" logic is used; as long as one of the conditions is met, the balance is considered complete.

[0085] Absolute error formula:

[0086] E abs =AB;

[0087] In the formula, E abs This is the absolute error;

[0088] A is the correct value;

[0089] B is the checksum (the value used for verification after user operation).

[0090] Relative error formula:

[0091] ;

[0092] Among them, E rel This is relative error;

[0093] It is understandable that by generating lever verification physical parameters through operation and comparing them with the generated lever verification physical parameters under set conditions, the verification physical parameters can be sent to the verification operation terminal at the beginning for verification.

[0094] Alternatively, the verification operation terminal can generate the lever verification physical parameters and send them to the verification server, which will then complete the verification.

[0095] It is understandable that the set requirements can be set according to the specific animation. For example, in the balance scale scene, it means achieving balance; in the crowbar scene, it means prying the object to achieve the set conditions; and in the pulley scene, it means lifting the object to achieve the set conditions.

[0096] To further prevent the machine from passing the above verification, in a further preferred embodiment, the failure to meet the set requirements specifically refers to the inability to complete the balance adjustment and / or failure to meet any one or more of the real-person verification conditions:

[0097] The conditions for real-person verification include: the number of operations exceeds a set number;

[0098] The operation time exceeded the set time;

[0099] The operation trajectory does not conform to the requirements of a real person's operation trajectory.

[0100] In a further preferred embodiment, the determination that the set requirement has not been met includes the following steps:

[0101] S41: Set the conditions for real-person verification, and assign a weight number Wi to the conditions for real-person verification;

[0102] The requirements for real-person verification include:

[0103] Extracted from trajectory data: Number of valid operations N = 4 (the user dragged and adjusted the position 4 times in total);

[0104] Total operation time T = 3200 milliseconds (total time from when the verification code is loaded to when the user clicks "verify");

[0105] Instantaneous velocity sequence of trajectory: Based on the timestamps and coordinates of trajectory points, calculate the instantaneous velocity (unit: px / s) between adjacent points to obtain the sequence {V i Assuming the calculated average speed... = 120 px / s, velocity variance σ² = 25 (px / s)².

[0106] Set the operation frequency verification weight W1 = 0.20;

[0107] The operation time verification weight W2 = 0.20;

[0108] The operation trajectory verification weight W3 = 0.60;

[0109] S42: Obtain values ​​that meet the verification conditions based on user usage characteristics;

[0110] Map each verification result to a binary score S. J ∈{0,1}, where 1 indicates passing (meets human characteristics) and 0 indicates failing.

[0111] Operation frequency verification: Let N be the number of valid user operations, and let N be the preset threshold. max (e.g., 5 times) ;

[0112] Operation time consumption verification: Let the total task time be t, and the reasonable manual operation interval be [T]. min ,T max (e.g., from 1 second to 60 seconds). ;

[0113] Operation trajectory analysis:

[0114] Calculate the average velocity of the instantaneous velocity sequence {Vi} of the trajectory. With variance ;

[0115] Let the lower bound of the variance of the speed of natural human operations be 6 var (e.g.) 10(px / s ) 2 The reasonable range for average speed is [V]. min V max ]. ;

[0116] Velocity variance: In this model, it is a necessary feature used to filter out trajectories with sufficient velocity fluctuations, and then to filter them for reasonableness through the average velocity range, thereby effectively distinguishing between human and machine behavior.

[0117] S43: Multiply the values ​​of different conditions by their corresponding weights and then sum them;

[0118] The weighted decision function and the comprehensive judgment score F are derived by weighted summation of the scores of each indicator: ;

[0119] Set the final decision threshold Θ (e.g., = Θ0.6):

[0120] S44: Compare the summed judgment value with the set judgment threshold to determine whether the set requirements are met.

[0121] ;

[0122] It is understood that the verification step can be performed by a verification server or a verification operation terminal. In this embodiment, see [link to documentation]. Figure 2 The verification operation terminal sends the parameters to the verification server, which then verifies the parameters and returns the verification result to the verification operation terminal.

[0123] Finally, the verification result is displayed on the verification operation terminal.

[0124] Example 2

[0125] A human-computer verification system based on the principle of physical lever balance, see [link to relevant documentation]. Figure 4 This includes verification servers and verification operation terminals;

[0126] The verification server is used to generate a set of lever physical parameters and send the lever physical parameters to the verification operation terminal;

[0127] For example, the verification server includes a verification code generation module and a verification code verification module, wherein the verification code verification module includes a lever balance determination unit and a human-machine behavior determination unit;

[0128] The verification code generation module is used to generate verification code parameters; the lever balance determination unit is used to verify the balance torque; and the human-machine behavior determination unit is used for real-person verification.

[0129] The verification terminal is used to render and generate a leverage verification code animation, allowing the customer to interact with the animation and determine whether the customer has passed the verification.

[0130] For example, the verification operation terminal includes a lever balance verification code rendering module and a rider motion trajectory point acquisition module;

[0131] The balanced CAPTCHA rendering module is used to render and generate a lever CAPTCHA animation.

[0132] The rider movement trajectory point acquisition module is used to collect user operations and rider trajectory.

[0133] The interaction relationships and steps between the modules are as described in Example 1 and... Figure 3 , Figure 4 As shown, it will not be elaborated further here.

[0134] Although the invention has been described herein with reference to illustrative embodiments, it should be understood that many other modifications and implementations can be devised by those skilled in the art, which will fall within the scope and spirit of the principles disclosed herein. More specifically, various variations and modifications can be made to the components and / or layout of the subject matter combination within the scope of this disclosure. Besides variations and modifications to the components and / or layout, other uses will be apparent to those skilled in the art.

Claims

1. A human-computer verification method based on the principle of physical lever balance, characterized in that, Includes the following steps: The verification server randomly generates a set of lever physical parameters and sends the lever physical parameters to the verification operation terminal; After receiving the physical parameters of the lever, the verification operation terminal renders and generates a lever verification code animation. Users interact with the lever verification code animation on the verification terminal; If the customer's actions during the verification code animation meet the set requirements, the verification will pass. If the set requirements are not met, the verification will be deemed a failure.

2. The human-computer verification method based on the physical lever balance principle as described in claim 1, characterized in that, The lever's physical parameters include one of the following: Option 1: The physical parameters of the lever include the mass of the item on the left, the mass of the item on the right, the mass of the sliding rider, the length of the left lever arm, the length of the right lever arm, and the net torque. Option 2: The physical parameters of the lever include the force applied on the left, the force applied on the right, the length of the left lever arm, the length of the right lever arm, and the net torque.

3. The human-computer verification method based on the physical lever balance principle as described in claim 2, characterized in that, In Option 1, the generated animation includes a balance scale animation; In Option 2, the generated animations include crowbar animation, seesaw animation, and pulley animation.

4. A human-computer verification method based on the principle of physical lever balance as described in claim 2 or 3, characterized in that, The specific requirements for the customer's operation of the leverage verification code animation are as follows: In Scheme 1, at least one of the physical parameters used in the lever verification code animation rendered by the verification operation terminal is an unbalanced parameter. The user adjusts the unbalanced parameter to make the balance lever reach a balanced state, thus meeting the set requirements. In Scheme 2, at least one of the physical parameters used in the lever verification code animation rendered by the verification operation terminal is a non-set parameter. The user adjusts the non-set parameter to make the crowbar, seesaw, and pulley lever meet the set requirements.

5. The human-computer verification method based on the physical lever balance principle as described in claim 4, characterized in that, In the lever physical parameters of Scheme 1, the net torque is a non-equilibrium parameter; The net torque is expressed and adjusted using the position of the balance rider.

6. The human-computer verification method based on the physical lever balance principle as described in claim 1, characterized in that, The user interacts with the lever verification code animation on the verification terminal, which specifically includes the following steps; Users click and / or swipe on elements in the interface; Based on the user's actions, the system generates physical parameters for lever verification and further renders the animation based on these parameters to generate user feedback.

7. The human-computer verification method based on the principle of physical lever balance as described in claim 1, characterized in that, The failure to meet the set requirements specifically refers to the inability to complete the balance adjustment and / or failure to meet any one or more of the real-person verification conditions: The conditions for real-person verification include: the number of operations exceeds a set number; The operation time exceeded the set time; The operation trajectory does not conform to the requirements of a real person's operation trajectory.

8. The human-computer verification method based on the physical lever balance principle as described in claim 1, characterized in that, The determination that the set requirements have not been met includes the following steps: Set the criteria for real-person verification and assign a weight Wi to those criteria. Set the conditions and values ​​for real-person verification; The value that meets the verification conditions is obtained based on the user's usage characteristics; Sum the values ​​of different conditions after multiplying them by their corresponding weights; The summed judgment value is compared with the set judgment threshold to determine whether the set requirements are met.

9. A human-machine verification system based on the principle of physical lever balance, characterized in that, This includes verification servers and verification operation terminals; The verification server is used to generate a set of lever physical parameters and send the lever physical parameters to the verification operation terminal; The verification terminal is used to render and generate a leverage verification code animation, allowing customers to interact with the animation and determine whether the customer has passed the verification.

Images