Cluster analysis of kitchen cabinet operation posture based on OpenPose technology

With the help of OpenPose technology, the key data of the user's operating posture when using the kitchen cabinet during cooking are obtained. The second-order clustering method is used to cluster the obtained data, and the relationship between the characteristics of human operating behavior posture and height among different categories is analyzed, to provide a scientific basis for the design of high adaptability customization scheme of the kitchen cabinet table in the later stage. The clustering results show that the deviation angle of the parts beyond the health threshold is different under different task states. The average deviation angle of the neck was more than 45°. In Category 3, the average waist deviation angle was 3–9 times of the other categories, and 68.05% of the operation time exceeded the health threshold. Therefore, in the design of the kitchen cabinet, the sizes of the cleaning area and preparation area need to be adjusted according to the sight distance to ensure that users maintain a healthy operation behavior. In addition, it is found that the user sight distance range of 160–180 cm is [419,621] mm under the operation state, which can be used as a reference for design evaluation. •Using OpenPose to capture key skeletal points of the human body and cluster and analyze the main movements during cabinet operations.•Transformation of key points of the human skeleton into key joint pinch angles based on mathematical models.•The limb frequency distribution was combined with the operational comfort scale to obtain the comfort range of human cooking behavior.