Indoor automatic layout method based on sliding window features and regression prediction

Abstract

The invention provides an indoor automatic layout method based on sliding window characteristics and regression prediction, which comprises three parts of house type data characterization, model selection and training, house type layout and post-processing. First of all, the characteristics of house type boundary are extracted by sliding window. Secondly, the characteristics of door and window areextracted by sliding window. Thirdly, the wall features are extracted by sliding window. Finally, the ratio factor between the side length of the layout element and the side length of the house typeis calculated to obtain the input characteristics of the element. The output characteristics of the element are composed of the normalized upper left corner coordinates of the relative house type andthe horizontal and vertical direction categories. Using the random forest regression model for off-line training, the complex association rules and layout rules between layout elements are automatically extracted from a large number of designed household data, and the layout model is obtained, and the prediction results are fine-tuned including wall constraints, overlapping constraints and other rules, and the final layout results are obtained.

Description

technical field [0001] The invention relates to the field of machine learning and smart home decoration design, in particular to an indoor automatic layout method based on sliding window features and regression prediction. Background technique [0002] Traditional interior decoration design requires designers to design different house plans according to the characteristics of the house type and the needs of the owners, which involves a lot of repetitive work. Indoor automatic layout refers to the use of layout algorithms to give the optimal placement of elements according to the given apartment structure, elements to be arranged and other individual needs. As the core functional module in the field of Internet home decoration, indoor automatic layout has important practical significance. On the one hand, it can improve the work efficiency of designers, shorten the design time of floor plans, and reduce repetitive work; on the other hand, it can provide owners with online per...

AI Technical Summary

Problems solved by technology

This method is essentially a rule-based target optimization algorithm. Its disadvantages are: ① only consider the relationship between elements in the case base, and do not consider the relationship between elements and house types. In actual situations, the same case base may appear in different house types. different relationships, which will lead to non-optimal layout of the case base; ②Continuous iterative optimization is required in the layout process, which is time-consuming

On the one hand, the rule combination and classification of the furniture to be laid out will lead to the loss of individuality; on the other hand, the tabu search algorithm is also a kind of optimization algorithm, which needs to design a rule-based objective function, and requires a large The amount of calculation is relatively time-consuming

[0005] The existing layout methods are mainly based on rule constraints. The disadvantages of this type of method are: ① In order to cover as many layout elements as possible, a large number of rule constraints need to be designed; The rule constraints will lead to the lack of personalization

Images