CN111368762A - Robot gesture recognition method based on improved K-means clustering algorithm - Google Patents
Robot gesture recognition method based on improved K-means clustering algorithmDownload PDFInfo
- Publication number
- CN111368762A CN111368762ACN202010157400.9ACN202010157400ACN111368762ACN 111368762 ACN111368762 ACN 111368762ACN 202010157400 ACN202010157400 ACN 202010157400ACN 111368762 ACN111368762 ACN 111368762A
- Authority
- CN
- China
- Prior art keywords
- robot
- sample
- improved
- category
- cluster
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Images
Classifications
- G—PHYSICS
- G06—COMPUTING OR CALCULATING; COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V40/00—Recognition of biometric, human-related or animal-related patterns in image or video data
- G06V40/20—Movements or behaviour, e.g. gesture recognition
- G06V40/28—Recognition of hand or arm movements, e.g. recognition of deaf sign language
- G—PHYSICS
- G06—COMPUTING OR CALCULATING; COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/23—Clustering techniques
- G06F18/232—Non-hierarchical techniques
- G06F18/2321—Non-hierarchical techniques using statistics or function optimisation, e.g. modelling of probability density functions
- G06F18/23213—Non-hierarchical techniques using statistics or function optimisation, e.g. modelling of probability density functions with fixed number of clusters, e.g. K-means clustering
- G—PHYSICS
- G06—COMPUTING OR CALCULATING; COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
- G06F18/241—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches
- G06F18/2413—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches based on distances to training or reference patterns
- G06F18/24147—Distances to closest patterns, e.g. nearest neighbour classification
Landscapes
- Engineering & Computer Science (AREA)
- Data Mining & Analysis (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- Computer Vision & Pattern Recognition (AREA)
- General Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- Evolutionary Biology (AREA)
- Evolutionary Computation (AREA)
- Bioinformatics & Computational Biology (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Artificial Intelligence (AREA)
- Life Sciences & Earth Sciences (AREA)
- Probability & Statistics with Applications (AREA)
- Health & Medical Sciences (AREA)
- General Health & Medical Sciences (AREA)
- Psychiatry (AREA)
- Social Psychology (AREA)
- Human Computer Interaction (AREA)
- Multimedia (AREA)
- Image Analysis (AREA)
- Manipulator (AREA)
Abstract
Translated fromChinese
Description
Translated fromChinese技术领域technical field
本发明涉及机器人手势识别领域，特别是涉及基于改进的K-means聚类算法的机器人手势识别方法。The invention relates to the field of robot gesture recognition, in particular to a robot gesture recognition method based on an improved K-means clustering algorithm.
背景技术Background technique
随着人工智能和虚拟现实技术的不断发展，人机交互系统已经成为当前的研究热点。现今，作为一种新兴的人机交互方式，手势识别得到了很多研究者的重视，并产生了一系列有效的成果，且在诸如智能机器人、智能驾驶等设备中得到了广泛的应用。手势识别，简单的来说就是让机器在视觉或传感器采集系统的辅助下来理解人类所想要表达的思想，即通过无接触的方式完成交互过程，从而通过机器人完成相应的动作，在真正意义上实现智能化。With the continuous development of artificial intelligence and virtual reality technology, human-computer interaction system has become a current research hotspot. Nowadays, as an emerging human-computer interaction method, gesture recognition has attracted the attention of many researchers, and has produced a series of effective results, and has been widely used in devices such as intelligent robots and intelligent driving. Gesture recognition, in simple terms, is to let the machine understand the thoughts that humans want to express with the help of the vision or sensor acquisition system, that is, to complete the interaction process in a non-contact way, so as to complete the corresponding actions through the robot, in the true sense Realize intelligence.
针对机器人手势识别的问题，国内涉及该问题解决方案的专利有“一种基于深度视觉的协作机器人手势识别方法及装置”(201910176271.5)，预先获取手势模板集合，同时获取待识别手势的若干张深度图；针对每一个手势模板，获取所述待识别手势与所述手势模板的距离，并将与所述待识别手势之间距离最小的手势模板作为所述待识别手势的识别结果，进而根据对应于识别结果的控制参数进行协作机器人的控制。国家发明专利“一种基于智能机器人的手势识别方法”(201910118356.8)，通过调用机器人自带的摄像头获取手势图像，并建立手势模板；通过基于肤色检测和基于最大类间方差对手势进行分割；使用中值滤波算法对分割后的手势图像进行去噪，并提取手势边缘轮廓；而后基于手势模板、手势边缘轮廓，采用欧式距离的模板匹配法，得到识别结果。以上两个发明专利都是对手势图片进行识别，过大维度的图片数据，一方面增大了模型的训练难度，另一方面在实际应用时，也增加了模型判别的时间。Aiming at the problem of robot gesture recognition, the domestic patent related to the solution to this problem is "A method and device for collaborative robot gesture recognition based on depth vision" (201910176271.5), which pre-obtains a set of gesture templates and simultaneously acquires several depths of the gesture to be recognized. Figure; For each gesture template, obtain the distance between the gesture to be recognized and the gesture template, and use the gesture template with the smallest distance between the gesture to be recognized as the recognition result of the gesture to be recognized, and then according to the corresponding Control the collaborative robot based on the control parameters of the recognition result. The national invention patent "A Gesture Recognition Method Based on Intelligent Robot" (201910118356.8), obtains the gesture image by calling the camera that comes with the robot, and establishes the gesture template; the gesture is segmented based on skin color detection and maximum inter-class variance; using The median filter algorithm denoises the segmented gesture image and extracts the gesture edge contour; then, based on the gesture template and the gesture edge contour, the Euclidean distance template matching method is used to obtain the recognition result. The above two invention patents both recognize gesture pictures. Too large-dimensional picture data increases the difficulty of model training on the one hand, and increases the time for model discrimination in practical applications.
发明内容SUMMARY OF THE INVENTION
为解决上述问题，本发明在微纳光纤传感器、K-means聚类算法的基础上，提出了基于改进的K-means聚类算法的机器人手势识别方法。首先利用微纳光纤传感器采集不用手势下对应的数据；而后利用改进的K-means聚类算法依次确定不同手势对应的聚类中心和类别阈值；同时该模型支持在线的更新优化，大大的提高了模型的泛化性；最后成功应用于实际，实现了机器人对不同手势的精准识别。为达此目的，本发明提供基于改进的K-means聚类算法的机器人手势识别方法，具体步骤如下，其特征在于：In order to solve the above problems, the present invention proposes a robot gesture recognition method based on the improved K-means clustering algorithm on the basis of the micro-nano optical fiber sensor and the K-means clustering algorithm. First, the micro-nano optical fiber sensor is used to collect the corresponding data without gestures; then the improved K-means clustering algorithm is used to sequentially determine the cluster centers and category thresholds corresponding to different gestures; at the same time, the model supports online update optimization, which greatly improves the The generalization of the model; finally, it was successfully applied in practice, realizing the precise recognition of different gestures by the robot. To achieve this purpose, the present invention provides a robot gesture recognition method based on an improved K-means clustering algorithm, the specific steps are as follows, and it is characterized in that:
步骤1，利用嵌有微纳光纤传感器的手套采集手部运动的数据，其中传感器采集数据维度为6维；Step 1, using a glove embedded with a micro-nano fiber optic sensor to collect hand motion data, wherein the dimension of the data collected by the sensor is 6 dimensions;
步骤2，通过手套上的WIFI模块将微纳光纤传感器采集到的数据上传给机器人；Step 2, upload the data collected by the micro-nano optical fiber sensor to the robot through the WIFI module on the glove;
步骤3，机器人结合改进的K-means聚类算法预先确定不同手势对应的聚类中心；Step 3, the robot pre-determines the cluster centers corresponding to different gestures in combination with the improved K-means clustering algorithm;
步骤4，计算当前微纳传感器采集的数据至预先确定的不同手势的聚类中心的欧氏距离；Step 4, calculating the Euclidean distance from the data collected by the current micro-nano sensor to the predetermined cluster centers of different gestures;
步骤5，将所计算的各欧氏距离与相应类别的阈值进行比较，如果欧式距离比类别阈值低，则判定为该类，否则，重新训练模型；Step 5, compare the calculated Euclidean distance with the threshold of the corresponding category, if the Euclidean distance is lower than the category threshold, it is determined as this category, otherwise, the model is retrained;
步骤6，机器人根据判断的结果完成相应的动作，至此，一个完整的闭环结束。Step 6, the robot completes the corresponding action according to the judgment result, so far, a complete closed loop ends.
进一步，步骤3中利用改进的K-means聚类算法预先确定不同手势对应的聚类中心的具体步骤为：Further, in step 3, the specific steps of using the improved K-means clustering algorithm to predetermine the cluster centers corresponding to different gestures are:
步骤3.1，在所有的样本点中任意选择一个样本点作为第一个类别的初始聚类中心c1；Step 3.1, arbitrarily select a sample point from all the sample points as the initial cluster center c1 of the first category;
步骤3.2，对于整个训练样本集X＝{xj|j＝1,2,...,n}，计算每一个样本x至聚类中心的距离，将最大距离对应的样本的所在位置作为新的聚类中心；Step 3.2, for the entire training sample set X={xj |j=1,2,...,n}, calculate the distance from each sample x to the cluster center, and use the location of the sample corresponding to the maximum distance as the new the cluster center;
步骤3.3，重复步骤3.2，直至确定k个聚类中心ci(1≤i≤k)；Step 3.3, repeat step 3.2 until k cluster centers ci (1≤i≤k) are determined;
步骤3.4，对整个训练样本集X＝{xj|j＝1,2,...,n}，分别计算各样本点xj到步骤3.3所确定的k个聚类中心ci(1≤i≤k)的欧式距离，对于s维的样本来说，样本xj到第i类聚类中心ci的欧氏距离为：Step 3.4, for the entire training sample set X={xj |j=1,2,...,n}, respectively calculate each sample point xj to the k cluster centers ci determined in step 3.3 (1≤ The Euclidean distance of i≤k), for the s-dimensional sample, the Euclidean distance from the sample xj to the i-th cluster center ci is:
步骤3.5，将样本分类到最近的欧式距离所在的类别中，遍历整个样本空间后完成对k个类簇的构建；Step 3.5, classify the samples into the category of the nearest Euclidean distance, and complete the construction of k clusters after traversing the entire sample space;
步骤3.6，对每个类簇，以该簇内所有样本点的均值向量作为新的类簇中心，即类簇的更新准则为：Step 3.6, for each cluster, the mean vector of all sample points in the cluster is used as the new cluster center, that is, the update criterion of the cluster is:
式中，ci为更新后的各类簇的中心，mi表示第i个类簇中样本的总数，
步骤3.7，重复步骤3.4-步骤3.6，直至平方误差函数收敛或迭代次数达到设定的次数，其中平方误差函数表达式为：Step 3.7, repeat steps 3.4-3.6 until the squared error function converges or the number of iterations reaches the set number of times, where the squared error function expression is:
进一步，步骤5中若实时采集的数据至各聚类中心的欧氏距离比任意的类别阈值大，则重新训练模型的具体描述为：Further, in step 5, if the Euclidean distance from the real-time collected data to each cluster center is larger than an arbitrary category threshold, the specific description of the retraining model is:
通过先验知识给实时采集的样本制作标签，然后将该数据带入到已训练完成的模型中对模型进行更新修正：更新各类别的聚类中心和相应的类别阈值。此处模型支持更新优化，大大的提高了模型的泛化性。Label the samples collected in real time through prior knowledge, and then bring the data into the trained model to update and correct the model: update the cluster centers of each category and the corresponding category thresholds. The model here supports update optimization, which greatly improves the generalization of the model.
本发明基于改进的K-means聚类算法的机器人手势识别方法，有益效果：本发明的技术效果在于：The present invention is based on the improved K-means clustering algorithm robot gesture recognition method, beneficial effects: the technical effect of the present invention is:
1.本发明利用六维微纳光纤传感器实时采集当前手势下的数据，较传统的通过图片形式进行手势识别的数据采集方法有着更低的样本维度，缩短了模型的训练时间，提高了模型的判别速度，同时也保证了很高的精度；1. The present invention uses a six-dimensional micro-nano optical fiber sensor to collect data under the current gesture in real time, which has a lower sample dimension than the traditional data collection method of gesture recognition through the form of pictures, shortens the training time of the model, and improves the performance of the model. Discriminating speed, but also ensuring high accuracy;
2.本发明利用改进的K-means聚类算法对不同的手势进行聚类分析，较传统的K-means聚类方法在避免出现错误的初始聚类中心方面有着更好的改善，同时可以精确的实现对不同手势的分类以及类别阈值的确定；2. The present invention utilizes the improved K-means clustering algorithm to perform cluster analysis on different gestures, which is better than the traditional K-means clustering method in avoiding wrong initial clustering centers, and can accurately The realization of the classification of different gestures and the determination of the category threshold;
3.本发明的机器人手势识别模型支持优化更新，即：当已训练好的模型无法对实时采集的数据进行分类判别，则将此时的数据作为训练数据代入至模型中进行重新训练，从而实现对各手势类别对应的类别中心和类别阈值的更新，因此大大的提高了模型的泛化性。3. The robot gesture recognition model of the present invention supports optimization and updating, that is: when the trained model cannot classify and discriminate the data collected in real time, the data at this time is substituted into the model as training data for retraining, thereby realizing The update of the category center and category threshold corresponding to each gesture category greatly improves the generalization of the model.
附图说明Description of drawings
图1为本发明的流程图；Fig. 1 is the flow chart of the present invention;
图2为本发明中利用改进的K-means聚类算法识别不同手势的聚类中心和与类别阈值的示意图；Fig. 2 utilizes improved K-means clustering algorithm in the present invention to identify the cluster centers of different gestures and the schematic diagram of category threshold;
具体实施方式Detailed ways
下面结合附图与具体实施方式对本发明作进一步详细描述：The present invention will be described in further detail below in conjunction with the accompanying drawings and specific embodiments:
本发明提出了基于改进的K-means聚类算法的机器人手势识别方法，旨在简单高效的实现机器人对不同手势的精确识别。The invention proposes a robot gesture recognition method based on an improved K-means clustering algorithm, aiming to realize the precise recognition of different gestures by a robot simply and efficiently.
图1为本发明的流程图。下面结合流程图对本发明的步骤作详细介绍。FIG. 1 is a flow chart of the present invention. The steps of the present invention will be described in detail below with reference to the flow chart.
步骤1，利用嵌有微纳光纤传感器的手套采集手部运动的数据，其中传感器采集数据维度为6维；Step 1, using a glove embedded with a micro-nano fiber optic sensor to collect hand motion data, wherein the dimension of the data collected by the sensor is 6 dimensions;
步骤2，通过手套上的WIFI模块将微纳光纤传感器采集到的数据上传给机器人；Step 2, upload the data collected by the micro-nano optical fiber sensor to the robot through the WIFI module on the glove;
步骤3，机器人结合改进的K-means聚类算法预先确定不同手势对应的聚类中心；Step 3, the robot pre-determines the cluster centers corresponding to different gestures in combination with the improved K-means clustering algorithm;
步骤3.1，在所有的样本点中任意选择一个样本点作为第一个类别的初始聚类中心c1；Step 3.1, arbitrarily select a sample point from all the sample points as the initial cluster center c1 of the first category;
步骤3.2，对于整个训练样本集X＝{xj|j＝1,2,...,n}，计算每一个样本x至聚类中心的距离，将最大距离对应的样本的所在位置作为新的聚类中心；Step 3.2, for the entire training sample set X={xj |j=1,2,...,n}, calculate the distance from each sample x to the cluster center, and use the location of the sample corresponding to the maximum distance as the new the cluster center;
步骤3.3，重复步骤3.2，直至确定k个聚类中心ci(1≤i≤k)；Step 3.3, repeat step 3.2 until k cluster centers ci (1≤i≤k) are determined;
步骤3.4，对整个训练样本集X＝{xj|j＝1,2,...,n}，分别计算各样本点xj到步骤3.3所确定的k个聚类中心ci(1≤i≤k)的欧式距离，对于s维的样本来说，样本xj到第i类聚类中心ci的欧氏距离为：Step 3.4, for the entire training sample set X={xj |j=1,2,...,n}, respectively calculate each sample point xj to the k cluster centers ci determined in step 3.3 (1≤ The Euclidean distance of i≤k), for the s-dimensional sample, the Euclidean distance from the sample xj to the i-th cluster center ci is:
步骤3.5，将样本分类到最近的欧式距离所在的类别中，遍历整个样本空间后完成对k个类簇的构建；Step 3.5, classify the samples into the category of the nearest Euclidean distance, and complete the construction of k clusters after traversing the entire sample space;
步骤3.6，对每个类簇，以该簇内所有样本点的均值向量作为新的类簇中心，即类簇的更新准则为：Step 3.6, for each cluster, the mean vector of all sample points in the cluster is used as the new cluster center, that is, the update criterion of the cluster is:
式中，ci为更新后的各类簇的中心，mi表示第i个类簇中样本的总数，
步骤3.7，重复步骤3.4-步骤3.6，直至平方误差函数收敛或迭代次数达到设定的次数，其中平方误差函数表达式为：Step 3.7, repeat steps 3.4-3.6 until the squared error function converges or the number of iterations reaches the set number of times, where the squared error function expression is:
步骤4，计算当前微纳传感器采集的数据至预先确定的不同手势的聚类中心的欧氏距离；Step 4, calculating the Euclidean distance from the data collected by the current micro-nano sensor to the predetermined cluster centers of different gestures;
步骤5，将所计算的各欧氏距离与相应类别的阈值进行比较，如果欧式距离比类别阈值低，则判定为该类，否则，重新训练模型。其中重新训练模型的具体描述为：通过先验知识给实时采集的样本制作标签，然后将该数据带入到已训练完成的模型中对模型进行更新修正：更新各类别的聚类中心和相应的类别阈值。此处模型支持更新优化，大大的提高了模型的泛化性。Step 5: Compare the calculated Euclidean distances with the thresholds of the corresponding categories. If the Euclidean distances are lower than the category thresholds, the category is determined. Otherwise, the model is retrained. The specific description of the retraining model is: label the samples collected in real time through prior knowledge, and then bring the data into the trained model to update and correct the model: update the cluster centers of each category and the corresponding Category threshold. The model here supports update optimization, which greatly improves the generalization of the model.
步骤6，机器人根据判断的结果完成相应的动作，至此，一个完整的闭环结束。Step 6, the robot completes the corresponding action according to the judgment result, so far, a complete closed loop ends.
图2为本发明中利用改进的K-means聚类算法识别不同类别的聚类中心和与类别阈值的示意图。从图示可以看出，利用改进的K-means聚类算法可以简单有效的确定不同类别的聚类中心和边界，进而得出相应的类别阈值。FIG. 2 is a schematic diagram of identifying different types of cluster centers and thresholds by using the improved K-means clustering algorithm in the present invention. It can be seen from the figure that the improved K-means clustering algorithm can simply and effectively determine the cluster centers and boundaries of different categories, and then obtain the corresponding category thresholds.
以上所述，仅是本发明的较佳实施例而已，并非是对本发明作任何其他形式的限制，而依据本发明的技术实质所作的任何修改或等同变化，仍属于本发明所要求保护的范围。The above are only preferred embodiments of the present invention, and are not intended to limit the present invention in any other form, and any modifications or equivalent changes made according to the technical essence of the present invention still fall within the scope of protection of the present invention. .
Claims (3)
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010157400.9ACN111368762A (en) | 2020-03-09 | 2020-03-09 | Robot gesture recognition method based on improved K-means clustering algorithm |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010157400.9ACN111368762A (en) | 2020-03-09 | 2020-03-09 | Robot gesture recognition method based on improved K-means clustering algorithm |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN111368762Atrue CN111368762A (en) | 2020-07-03 |
Family
ID=71206751
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202010157400.9APendingCN111368762A (en) | 2020-03-09 | 2020-03-09 | Robot gesture recognition method based on improved K-means clustering algorithm |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN111368762A (en) |
Cited By (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112016621A (en)* | 2020-08-28 | 2020-12-01 | 上海第一财经数据科技有限公司 | Training method of classification model, color classification method and electronic equipment |
| CN112035663A (en)* | 2020-08-28 | 2020-12-04 | 京东数字科技控股股份有限公司 | Cluster analysis method, device, equipment and storage medium |
| CN112446296A (en)* | 2020-10-30 | 2021-03-05 | 杭州易现先进科技有限公司 | Gesture recognition method and device, electronic device and storage medium |
| CN113516063A (en)* | 2021-06-29 | 2021-10-19 | 北京精密机电控制设备研究所 | Motion mode identification method based on K-Means and gait cycle similarity |
| CN113608074A (en)* | 2021-06-17 | 2021-11-05 | 国网浙江省电力有限公司营销服务中心 | Automatic online monitoring method and system for multi-epitope voltage withstand test device |
Citations (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105571045A (en)* | 2014-10-10 | 2016-05-11 | 青岛海尔空调电子有限公司 | Somatosensory identification method, apparatus and air conditioner controller |
| CN105956604A (en)* | 2016-04-20 | 2016-09-21 | 广东顺德中山大学卡内基梅隆大学国际联合研究院 | Action identification method based on two layers of space-time neighborhood characteristics |
| CN106845348A (en)* | 2016-12-20 | 2017-06-13 | 南京信息工程大学 | A kind of gesture identification method based on arm surface electromyographic signal |
| CN107014411A (en)* | 2017-04-05 | 2017-08-04 | 浙江大学 | A kind of flexible micro-nano fiber angle sensor chip and sensor and preparation method |
| CN108983973A (en)* | 2018-07-03 | 2018-12-11 | 东南大学 | A kind of humanoid dexterous myoelectric prosthetic hand control method based on gesture identification |
| CN109032337A (en)* | 2018-06-28 | 2018-12-18 | 济南大学 | A kind of KEM Gesture Recognition Algorithm based on data glove |
| CN109547136A (en)* | 2019-01-28 | 2019-03-29 | 北京邮电大学 | Distributed collaborative frequency spectrum sensing method based on minimax apart from sub-clustering |
| CN110147754A (en)* | 2019-05-17 | 2019-08-20 | 金陵科技学院 | A kind of dynamic gesture identification method based on VR technology |
| CN110348323A (en)* | 2019-06-19 | 2019-10-18 | 广东工业大学 | A kind of wearable device gesture identification method based on Neural Network Optimization |
- 2020
- 2020-03-09CNCN202010157400.9Apatent/CN111368762A/enactivePending
Patent Citations (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105571045A (en)* | 2014-10-10 | 2016-05-11 | 青岛海尔空调电子有限公司 | Somatosensory identification method, apparatus and air conditioner controller |
| CN105956604A (en)* | 2016-04-20 | 2016-09-21 | 广东顺德中山大学卡内基梅隆大学国际联合研究院 | Action identification method based on two layers of space-time neighborhood characteristics |
| CN106845348A (en)* | 2016-12-20 | 2017-06-13 | 南京信息工程大学 | A kind of gesture identification method based on arm surface electromyographic signal |
| CN107014411A (en)* | 2017-04-05 | 2017-08-04 | 浙江大学 | A kind of flexible micro-nano fiber angle sensor chip and sensor and preparation method |
| CN109032337A (en)* | 2018-06-28 | 2018-12-18 | 济南大学 | A kind of KEM Gesture Recognition Algorithm based on data glove |
| CN108983973A (en)* | 2018-07-03 | 2018-12-11 | 东南大学 | A kind of humanoid dexterous myoelectric prosthetic hand control method based on gesture identification |
| CN109547136A (en)* | 2019-01-28 | 2019-03-29 | 北京邮电大学 | Distributed collaborative frequency spectrum sensing method based on minimax apart from sub-clustering |
| CN110147754A (en)* | 2019-05-17 | 2019-08-20 | 金陵科技学院 | A kind of dynamic gesture identification method based on VR technology |
| CN110348323A (en)* | 2019-06-19 | 2019-10-18 | 广东工业大学 | A kind of wearable device gesture identification method based on Neural Network Optimization |
Non-Patent Citations (1)
| Title |
|---|
| 王道东: "基于人机交互系统的手势识别方法研究", 《中国优秀硕士学位论文全文数据库 信息科技辑》* |
Cited By (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112016621A (en)* | 2020-08-28 | 2020-12-01 | 上海第一财经数据科技有限公司 | Training method of classification model, color classification method and electronic equipment |
| CN112035663A (en)* | 2020-08-28 | 2020-12-04 | 京东数字科技控股股份有限公司 | Cluster analysis method, device, equipment and storage medium |
| CN112016621B (en)* | 2020-08-28 | 2023-11-24 | 上海应帆数字科技有限公司 | Training method of classification model, color classification method and electronic equipment |
| CN112035663B (en)* | 2020-08-28 | 2024-05-17 | 京东科技控股股份有限公司 | Cluster analysis method, device, equipment and storage medium |
| CN112446296A (en)* | 2020-10-30 | 2021-03-05 | 杭州易现先进科技有限公司 | Gesture recognition method and device, electronic device and storage medium |
| CN112446296B (en)* | 2020-10-30 | 2024-10-22 | 杭州易现先进科技有限公司 | Gesture recognition method and device, electronic device and storage medium |
| CN113608074A (en)* | 2021-06-17 | 2021-11-05 | 国网浙江省电力有限公司营销服务中心 | Automatic online monitoring method and system for multi-epitope voltage withstand test device |
| CN113608074B (en)* | 2021-06-17 | 2024-05-28 | 国网浙江省电力有限公司营销服务中心 | Automatic online monitoring method and system for multi-epitope withstand voltage test device |
| CN113516063A (en)* | 2021-06-29 | 2021-10-19 | 北京精密机电控制设备研究所 | Motion mode identification method based on K-Means and gait cycle similarity |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN111368762A (en) | Robot gesture recognition method based on improved K-means clustering algorithm | |
| CN109858406B (en) | Key frame extraction method based on joint point information | |
| CN104463100B (en) | Intelligent wheel chair man-machine interactive system and method based on human facial expression recognition pattern | |
| CN110728694B (en) | Long-time visual target tracking method based on continuous learning | |
| CN103268495B (en) | Human body behavior modeling recognition methods based on priori knowledge cluster in computer system | |
| CN105740823B (en) | Dynamic gesture track recognizing method based on depth convolutional neural networks | |
| CN106407958B (en) | Face feature detection method based on double-layer cascade | |
| CN108921107A (en) | Pedestrian's recognition methods again based on sequence loss and Siamese network | |
| Kaâniche et al. | Recognizing gestures by learning local motion signatures of HOG descriptors | |
| CN109522853A (en) | Face datection and searching method towards monitor video | |
| CN109543615B (en) | A dual-learning model target tracking method based on multi-level features | |
| CN109753897B (en) | Behavior recognition method based on memory cell reinforcement-time sequence dynamic learning | |
| CN112084898B (en) | Assembly operation action recognition method based on static and dynamic separation | |
| CN106778501A (en) | Video human face ONLINE RECOGNITION method based on compression tracking with IHDR incremental learnings | |
| CN112883922B (en) | Sign language identification method based on CNN-BiGRU neural network fusion | |
| CN112818175A (en) | Factory worker searching method and training method of worker recognition model | |
| CN109087337B (en) | Long-time target tracking method and system based on hierarchical convolution characteristics | |
| CN112381047B (en) | Enhanced recognition method for facial expression image | |
| CN105976397B (en) | A kind of method for tracking target | |
| CN113129336A (en) | End-to-end multi-vehicle tracking method, system and computer readable medium | |
| CN108229401A (en) | A kind of multi-modal Modulation recognition method based on AFSA-SVM | |
| CN110880010A (en) | Visual SLAM closed loop detection algorithm based on convolutional neural network | |
| CN108898623A (en) | Method for tracking target and equipment | |
| CN101799875A (en) | Target detection method | |
| CN108830222A (en) | A kind of micro- expression recognition method based on informedness and representative Active Learning |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| RJ01 | Rejection of invention patent application after publication | Application publication date:20200703 | |
| RJ01 | Rejection of invention patent application after publication |