Intelligent cockpit gesture recognition and analysis device

Abstract

The utility model relates to the technical field of gesture recognition, and discloses an intelligent cockpit gesture recognition and analysis device which comprises a mechanical arm, a data assembly and an ambient light adjusting unit. The data component comprises a node recognizer and a data comparator; the node recognizer is provided with a recognition end and a data end, the recognition end is arranged on the mechanical arm, and the data end is connected with one end of the data comparator; the other end of the data comparator is connected with a result output end of the gesture recognition system; the device is connected with a gesture recognition system of the intelligent cockpit, the light environment in the automobile cockpit is changed through the environment light adjusting unit, the recognition accuracy of automobile gesture recognition in different light environments is detected, the mode of the mechanical arm and the node recognizer serves as the comparison standard of automobile gesture recognition, and the recognition accuracy of the automobile gesture recognition is improved. The recognition results of the device and an automobile gesture recognition system are compared through the data comparator, the practicability of the device is improved, and the comprehensiveness and accuracy of gesture recognition detection analysis are improved.

Description

Technical Field

[0001] The utility model relates to the technical field of gesture recognition, and in particular to an intelligent cockpit gesture recognition and analysis device. Background Art

[0002] As the level of automobile intelligence continues to improve, autonomous driving functions are also gradually being promoted and used, including gesture recognition functions. By installing a gesture recognition module in the smart cockpit, the driver's gesture data is obtained to control some functions in the car.

[0003] However, there are some problems with this type of module in actual use. For example, in dim light, the camera recognition function cannot capture gestures; there is a gap between the gestures and the gestures in the comparison database, and there are multiple similar functions for one gesture. These factors will cause the gesture recognition module in the cockpit to not operate normally. How to improve the level of gesture recognition function has become a hot research direction.

[0004] After analysis, it is found that in addition to the performance impact of the gesture recognition module's own components, another important factor is the test method for evaluating the conventional gesture recognition function. Problems in the test process will also affect the accurate application of the final gesture recognition function. In the conventional test process, in order to ensure the comparability of the results, a unified light environment is usually used for testing. However, when testing the performance of different types of gesture recognition modules in a single environment, the results are limited. In addition, the flexibility of the test device structure adjustment is not high, and the applicable environment is single, which cannot meet the existing test requirements for adjusting the light environment. Utility Model Content

[0005] The utility model aims to provide an intelligent cockpit gesture recognition and analysis device, which is used to solve the technical problem that the existing gesture recognition and analysis device has low structural adjustment flexibility and cannot meet the requirements of the existing test for adjusting the light environment.

[0006] The basic solution provided by the utility model is: a smart cockpit gesture recognition and analysis device, connected to the gesture recognition system of the smart cockpit, used to compare and analyze the recognition results of the gesture recognition system and output the comparison results;

[0007] The device includes a mechanical arm, a data component and an ambient light adjustment unit; the data component includes a node identifier and a data comparator; the node identifier has an identification end and a data end, wherein the identification end is arranged on the mechanical arm, and the data end is connected to one end of the data comparator; the other end of the data comparator is connected to the result output end of the gesture recognition system.

[0008] The working principle and advantages of the utility model are as follows: the device changes the light environment in the car cockpit through the ambient light adjustment unit, detects the recognition accuracy of car gesture recognition under different light environments, and uses the mechanical arm and node identifier as the comparison standard for car gesture recognition, and compares the recognition result of the device with the recognition result of the car gesture recognition system through the data comparator, thereby realizing the function of analysis and detection, which not only increases the practicability of the device, but also improves the comprehensiveness and accuracy of gesture recognition detection analysis. BRIEF DESCRIPTION OF THE DRAWINGS

[0009] Figure 1 A schematic diagram of the structure of a mechanical arm of an intelligent cockpit gesture recognition and analysis device provided in an embodiment of the utility model Figure 1 ;

[0010] Figure 2 A schematic diagram of the structure of a mechanical arm of an intelligent cockpit gesture recognition and analysis device provided in an embodiment of the utility model Figure 2 ;

[0011] Figure 3 A schematic diagram of the structure of a lighting assembly of an intelligent cockpit gesture recognition and analysis device provided by an embodiment of the utility model Figure 1 ;

[0012] Figure 4 A schematic diagram of the structure of a lighting assembly of an intelligent cockpit gesture recognition and analysis device provided by an embodiment of the utility model Figure 2 ;

[0013] Figure 5 A schematic diagram of the structure of an adjustment circle of an intelligent cockpit gesture recognition and analysis device provided by an embodiment of the utility model Figure 1 ;

[0014] Figure 6 A schematic diagram of the structure of an adjustment circle of an intelligent cockpit gesture recognition and analysis device provided by an embodiment of the utility model Figure 2 ;

[0015] Figure 7 A schematic structural diagram of a curtain assembly of a smart cockpit gesture recognition and analysis device provided in an embodiment of the utility model. DETAILED DESCRIPTION

[0016] The following is a further detailed description through specific implementation methods:

[0017] The marks in the drawings of the specification include: robotic arm 1, forearm 101, rear arm 102, hand structure 103, bidirectional rotating member 104, first rotating member 105, hinge structure 106, connecting rod 107, fixed platform 108, shaft structure 109, fixed component 2, fixed clamp 201, bidirectional sliding groove 202, lighting component 3, variable lamp 301, adjustable lampshade 302, adjustment ring 303, sliding groove 3031, inner groove 3032, extended edge 3033, arc-shaped connecting section 3034, guide rail 304, curtain component 4, sliding rail 401, slider 402, reel 403, fixed shaft 404, pull rod 405, node identifier 5, transmission line 501, marking point 502.

[0018] Embodiment 1

[0019] Basically as attached Figure 1 , Figure 3 and Figure 7 Shown: A smart cockpit gesture recognition and analysis device, connected to the gesture recognition system of the smart cockpit, is used to analyze the recognition results of the gesture recognition system.

[0020] The device includes a robotic arm 1, a data component and an ambient light adjustment unit; the data component includes a node identifier 5 and a data comparator; the node identifier 5 has an identification end and a data end, wherein the identification end is arranged on the robotic arm 1, and the data end is connected to one end of the data comparator; the other end of the data comparator is connected to the result output end of the gesture recognition system.

[0021] Specifically, Figure 1 and Figure 2 As shown, the robotic arm 1 includes a forearm 101, a rear arm 102 and a hand structure 103, and the forearm 101, the rear arm 102 and the hand structure 103 are all connected through a servo; the hand structure 103 is consistent with a real hand, and the hand structure 103 includes a palm and fingers, and the fingers are divided into three parts: front, middle and back, and the palm is divided into a front palm and a back palm to ensure the authenticity of the simulated gesture.

[0022] A first rotating member 105 is provided between the front arm 101 and the rear arm 102, and a hinge structure 106 is provided on one side of the first rotating member 105 (the inner side of the arm that can be bent), and connecting rods 107 are connected on both sides of the hinge structure 106, and the other ends of the connecting rods 107 on both sides are respectively connected to the front arm 101 and the rear arm 102, and fixed platforms 108 and rotating shaft structures 109 are provided at the connections. The middle of the fixed platform 108 is circular, so that the fixed platform 108 can be fixed on the main structure of the arm. When the connecting rod 107 and the hinge move relative to each other, the fixed platform 108 moves along the length direction of the arm, thereby realizing the movement of the arm.

[0023] One end of the forearm 101 is connected to the rear arm 102, and the other end of the forearm 101 is used as the connecting end of the robot arm 1, and is installed with a bidirectional rotating member 104, which includes an upper part and a lower part that are movably connected, so that the connecting end itself can achieve lateral rotation. A circular connecting port is provided at the upper part of the connecting end, and a groove is provided on the circular connecting port for installing the robot arm 1.

[0024] The device also includes a fixing component 2 for connecting to the robot arm 1, and the fixing component 2 is connected via a circular connection port at the connecting end of the robot arm 1. Corresponding grooves are provided on the circular connection port and the fixing component 2 so that the two can be connected and rotated at the same time.

[0025] The fixing assembly 2 includes a fixing clamp 201 and a base, and is connected to the robot arm 1 through the base. The base is provided with a bidirectional sliding groove 202, and the bidirectional sliding groove 202 intersects in a cross shape. The circular connection port at the connecting end of the robot arm 1 is movably connected to the bidirectional sliding groove 202 of the fixing assembly 2. In other embodiments, a sliding component may also be provided at the connecting end of the robot arm 1, and the sliding component is slidably connected to the bidirectional sliding groove 3031, and the movement of the robot arm 1 is achieved through this design.

[0026] When the analysis device is used inside a smart cockpit, the fixing component 2 is installed in the smart cockpit, and the robotic arm 1 can be installed in the smart cockpit, specifically at the seat, to simulate the arm position of the driver when sitting in the cockpit as realistically as possible, to ensure that the gesture data to be recognized is acquired normally and accurately, and to improve the accuracy of the recognition and analysis results.

[0027] Specifically, the data component includes a node identifier 5 and a data comparator. The node identifier 5 has an identification end and a data end, wherein the identification end is arranged on the robot arm 1, and the data end is connected to one end of the data comparator. The other end of the data comparator is connected to the gesture recognition system.

[0028] The node identifier 5 includes a data processor, a marking point 502, and a transmission line 501 connecting the data processor and the marking point 502. The transmission line 501 is connected in series with the marking point 502, and the end of the transmission line 501 is connected to the data processor. The identification end is the marking point 502, and the data end is the data processor. The specific material and composition of the marking point can refer to the marking point of the clay model in 3D scanning. The marking point 502 is provided with a marking point sticker, which is a circular sticker with a diameter of 6mm. A magnetic material is coated on the surface of the sticker and glue is on the back. The marking point sticker is attached to the joint point position of the robot arm 1 to collect the action information of the robot arm 1. The action information includes the moving trajectory and starting point of the marking point 502 on the robot arm 1, which is transmitted to the data processor through the transmission line 501 for identification and recording. The data processor uses the corresponding algorithm model of the moving trajectory and the starting and ending points. The Dijkstra algorithm can be used in combination with the built-in multiple gesture comparison standards to perform gesture recognition to ensure the accuracy of recognition and reduce the misrecognition rate of the analysis device. There can be many types of data processors, which can be a host composed of a CPU or a server software based on data processing on the market. It can be used as long as the function can be realized, and there is no limitation on the specific model.

[0029] The data comparator is provided with a sound component and a light component. The sound component is provided with two sounds, which respectively represent accurate recognition and incorrect recognition. The light component is composed of a warning light, which can emit red light and green light. The red light represents that the analysis result is accurate recognition, and the green light represents that the analysis result is incorrect recognition. The data processor is connected to one end of the data comparator to transmit the first recognition result of the data processor to the data comparator. The other end of the data comparator is connected to the gesture recognition system. After the robot arm 1 moves, the gesture recognition system simultaneously performs gesture recognition and outputs the second recognition result to the data comparator. The data comparator receives the first recognition result and the second recognition result, performs comparative analysis, and makes a comparison result. The comparison result is output in the form of sound and warning light to ensure the practicality of the device.

[0030] The standard gestures in the data processing terminal are consistent with the standard gesture data in the automobile gesture recognition system, ensuring the consistency of the basic standards for gesture recognition, reducing interference factors, and ensuring the effectiveness of comparative analysis.

[0031] Data Compare / NineData can be used as a data comparison platform. The data comparator and processor can be regarded as a computer. The data processing and comparison are based on the cloud server or local area network. This function can be completed. The model will not affect the implementation of the solution and is not limited. The data comparator is equipped with a memory and a USB output port, so that the comparison results in the data comparator can be copied and connected to the computer, which is convenient for staff to check.

[0032] Specifically, the ambient light adjustment unit includes a light assembly 3 and a plurality of curtain assemblies 4, which are used to change the light view in the comparative analysis space and detect the recognition accuracy of the automobile gesture recognition system under different light environments.

[0033] like Figure 3 and Figure 4 As shown, the lighting assembly 3 includes a variable lamp 301 and an adjustable lampshade 302. The variable lamp 301 is an intelligent lamp with adjustable light intensity, such as an LED lamp, as a first dimming method. The adjustable lampshade 302 includes a lampshade body, and a guide rail 304 and an adjustment ring 303 arranged on the outside of the lampshade body; a slide groove 3031 is arranged on the inner side of the adjustment ring 303, and the adjustment ring is movably connected to the lampshade body through the slide groove and the guide rail. The opening and closing degree of the lampshade body is adjusted by moving the adjustment ring 303 to adjust the light. When the lampshade is closed, the light refracted by the lampshade is incident light, and when the lampshade is separated, the light refracted by the lampshade is scattered light, which is a second dimming method.

[0034] like Figure 5 and Figure 6 As shown, the adjustment ring 303 includes N arc-shaped parts, each of which is provided with a slide groove 3031; the lampshade body includes N lampshade parts, each of which is trapezoidal in shape in the unfolded plan view, each of which is provided with a guide rail 304, and one lampshade part corresponds to one guide rail 304 and one slide groove 3031. N can be selected as 3, so that the adjustment is smoother. Specifically, the adjustment ring 303 is composed of a six-segment structure, among which the middle of three main bodies is provided with a slide groove 3031, which is adapted to the guide rail 304 and is used to move on the guide rail 304 along the length direction of the lampshade to realize the movable connection between the adjustment ring 303 and the lampshade body. Each section is provided with inner grooves 3032 on both sides, and an inward extending edge 3033 is provided at the opening of the inner grooves 3032. An arc-shaped connecting section 3034 is provided in the middle of the two adjacent sections. The arc-shaped connecting section 3034 is arranged in the inner grooves 3032 and slides between the two adjacent sections. Both ends of the arc-shaped connecting section 3034 are also provided with outward extending edges 3033, which are not fixed on the lampshade body, and act like a belt to connect the three sections of the main body; the lampshade is also divided into three sections, and the three sections of the lampshade are closed to form a truncated cone-shaped lampshade, but each individual structure has excess width on both sides, so that the connection between the three sections of the structure overlaps each other. This connection method ensures that there will be no gap when the lampshade is closed and released.

[0035] The main body of the lampshade is made of acrylic material, and the inner wall is affixed with a reflective film, so that the toughness of the main body of the lampshade can support the change of the shape of the lampshade. The lampshade is a truncated cone. When the adjustment ring 303 moves outward, the truncated cone-shaped lampshade center angle of the lampshade can be increased or decreased through the segmented structure, thereby changing the focusing effect of the lampshade. When the opening of the lampshade is the largest, the emitted light is scattered light, and when the opening of the lampshade is the smallest, the emitted light is concentrated light. The largest is trumpet-shaped, and the smallest is cylindrical. The smallest cylindrical shape is that the shape of the lampshade opening is limited by the structure of the adjustment ring 303, because the three sections of the adjustment ring 303 cannot overlap each other, and the smallest state of the adjustment ring 303 is to close into a ring. The minimum diameter of the ring of the adjustment ring 303 must be greater than or equal to the minimum diameter of the lampshade. In this way, the opening and closing degree of the lampshade is adjusted through the lampshade and the adjustment ring 303 to adjust the light.

[0036] The curtain 4 assembly includes a translucent curtain 4, two slide rails 401 and two sliders 402. The slide rail 401 is a multi-section slide rail 401. The sliders 402 are movably connected to the multi-section slide rails 401 in a one-to-one correspondence. The same side ends of the two slide rails 401 are provided with a reel 403. A fixed shaft 404 is provided between the two reels 403. A pull rod 405 is provided between the two sliders 402. One end of the translucent curtain 4 is fixed on the reel 403, and the other end of the translucent curtain 4 is fixed on the pull rod 405. The translucent curtain 4 is made of black flexible material. When in use, the slider 402 is moved to adjust different positions on the slide rail 401 to change the shading area of ​​the translucent curtain 4, which serves as a third dimming method. A set of curtain 4 assemblies can be installed on both sides of the cab windows and windshield, wherein the reel 403 is located at the top for easy adjustment.

[0037] The specific use process of this device is as follows:

[0038] S1, connect the equipment, stick the marking points of the node identifier 5 on the joints of the robot arm 1, connect the other end of the node identifier 5 to the data comparator, and connect the data comparator to the gesture recognition system of the car; the total number of all related nodes does not exceed 30, which meets the appropriate number of data collection points and obtains accurate gesture data.

[0039] S2, install the robot arm 1, fix the fixing component 2 on the car seat, connect the robot arm 1 with the fixing component 2 and ensure smooth sliding; check that all the marking points 502 of the robot arm 1 are connected, and power on the robot arm 1;

[0040] S3, under the preset light environment, the gesture recognition system is tested, the robot arm 1 is controlled to make various gestures, the moving trajectory and starting point of the mark point 502 on the robot arm 1 are transmitted to the data processor through the transmission line 501 for recognition and recording, the data processor performs gesture recognition by combining the moving trajectory and the starting and ending points with the built-in multiple gesture comparison standards, and outputs a first recognition result; the gesture recognition system performs gesture recognition at the same time, and outputs a second recognition result to the data comparator, the data comparator receives the first recognition result and the second recognition result, performs comparative analysis, and makes a comparison result, and outputs the comparison result in the form of sound and warning light.

[0041] There is no limit to the order of steps S1 and S2. There are many preset light environments in S3, including natural light isolation recognition, weak light recognition and strong light recognition. Each situation is divided into different light intensities. The three dimming methods achieved by variable light adjustment, lampshade adjustment and curtain movement are used independently or in combination to meet the test requirements of various lighting conditions. Through gesture recognition analysis under various lighting conditions, the gesture recognition system of the car is comprehensively analyzed based on the comprehensive recognition results.

[0042] The present embodiment provides an intelligent cockpit gesture recognition and analysis device, which changes the light environment of the test space through an ambient light adjustment unit to detect the recognition accuracy of the automobile gesture recognition system under different light environments; at the same time, a mechanical arm and a node identifier are used as comparison parties of the automobile gesture recognition system, and two recognition results are compared through a data comparator to achieve the function of analysis and detection; the device is simple, flexible to adjust, and highly adaptable, while increasing the practicality of the device, and can provide multi-dimensional results in the gesture recognition system test process, improve the result value of the system test, and be conducive to providing more data support for subsequent system optimization.

[0043] Embodiment 2

[0044] Different from the first embodiment, the warning light can emit light of multiple colors, including red, purple, green, blue, yellow and white light. The seven kinds of light are used to emit the gap between the vehicle recognition result and the device recognition result. From left to right, the gap gradually decreases. The warning light emits white light, indicating that the device recognition result is consistent with the recognition result of the vehicle recognition system.

[0045] The smart cockpit gesture recognition and analysis device provided in this embodiment makes the difference degree of the comparison recognition results clearer and more obvious, and can provide the tester with a more intuitive and real-time reminder of the result difference, so that the tester can choose to continue or pause the test according to the degree of difference, thereby balancing the effectiveness, accuracy and efficiency of the test.

[0046] The above is only an embodiment of the utility model. The common sense such as the known specific structure and characteristics in the scheme is not described in detail here. The ordinary technicians in the relevant field are aware of all the common technical knowledge in the technical field of the utility model before the application date or priority date, can obtain all the existing technologies in the field, and have the ability to apply the conventional experimental means before that date. The ordinary technicians in the relevant field can improve and implement the scheme in combination with their own abilities under the enlightenment given by this application. Some typical known structures or known methods should not become obstacles for ordinary technicians in the relevant field to implement this application. It should be pointed out that for the technicians in this field, without departing from the structure of the utility model, several deformations and improvements can be made, which should also be regarded as the protection scope of the utility model, and these will not affect the effect of the implementation of the utility model and the practicality of the patent.

Claims

1. An intelligent cockpit gesture recognition and analysis device, characterized in that: A gesture recognition system connected to the smart cockpit is used to compare and analyze the recognition results of the gesture recognition system and output the comparison results; The device includes a mechanical arm, a data component and an ambient light adjustment unit; the data component includes a node identifier and a data comparator; the node identifier has an identification end and a data end, wherein the identification end is arranged on the mechanical arm, and the data end is connected to one end of the data comparator; the other end of the data comparator is connected to the result output end of the gesture recognition system.

2. The intelligent cockpit gesture recognition and analysis device according to claim 1, characterized in that: The ambient light adjustment unit includes a lighting assembly; the lighting assembly includes a variable lamp and an adjustable lampshade; the adjustable lampshade includes a lampshade body, and a guide rail and an adjustment ring arranged on the outside of the lampshade body; a slide groove is provided on the inner side of the adjustment ring, and the adjustment ring is movably connected to the lampshade body through the slide groove and the guide rail.

3. The intelligent cockpit gesture recognition and analysis device according to claim 2, characterized in that: The adjustment ring includes N arc-shaped parts, each of which is provided with a slide groove; the lampshade body includes N lampshade splits, each of which is provided with a guide rail, and one lampshade split corresponds to one guide rail and one slide groove.

4. The intelligent cockpit gesture recognition and analysis device according to claim 1, characterized in that: The ambient light adjustment unit includes a curtain assembly; the curtain assembly includes a translucent curtain, two slide rails and two sliders, wherein the sliders are movably connected to the slide rails in a one-to-one correspondence, a reel is provided at the same side end of the two slide rails, a fixed shaft is provided between the two reels, and a pull rod is provided between the two sliders; one end of the translucent curtain is fixed on the reel, and the other end of the translucent curtain is fixed on the pull rod; the slide rail is a multi-section slide rail.

5. The intelligent cockpit gesture recognition and analysis device according to claim 1, characterized in that: The mechanical arm comprises a forearm, a rear arm and a hand structure, and the forearm, the rear arm and the hand structure are all connected through a steering gear.

6. The intelligent cockpit gesture recognition and analysis device according to claim 5, characterized in that: A first rotating member is provided between the front arm and the rear arm; a hinge structure is provided on one side of the first rotating member; connecting rods are connected to both sides of the hinge structure, and the other ends of the connecting rods on both sides are respectively connected to the front arm and the rear arm, and a fixed platform and a rotating shaft structure are provided at the connection points.

7. The intelligent cockpit gesture recognition and analysis device according to claim 5, characterized in that: One end of the forearm away from the rear arm serves as the connecting end of the robot arm, and is installed with a bidirectional rotating member; the bidirectional rotating member includes an upper part and a lower part that are movably connected; the upper part is provided with a circular connecting port for installing the robot arm.

8. The intelligent cockpit gesture recognition and analysis device according to claim 1, characterized in that: The node identifier includes a data processor, a marking point and a transmission line connecting the data processor and the marking point, wherein the identification end is the marking point and the data end is the data processor; the marking point is provided with a marking point sticker.

9. The intelligent cockpit gesture recognition and analysis device according to claim 1, characterized in that: The data comparator is provided with a sound component and a light component.

10. The intelligent cockpit gesture recognition and analysis device according to claim 1, characterized in that: It also includes a fixing component; the fixing component includes a connected fixing clamp and a base, and is connected to the mechanical arm through the base; the base is provided with a bidirectional sliding groove.

Images