Precise illegal building detection system

Abstract

The invention belongs to the technical field of image recognition target detection, and particularly relates to a precise illegal building detection system, which comprises a picture acquisition module for acquiring a picture of an illegal building according to a video of a pre-selected area; a labeling module for labeling the picture and obtaining a priori box according to the labeled picture; adivision module for dividing a data set formed by the picture into a training set and a test set; a preprocessing module for preprocessing the training set; a training module for training the YOLOV4 model according to the preprocessed training set; an adjusting module for adjusting the trained YOLOV4 model according to the test set; and a detection module for inputting the video into the adjustedYOLOV4 model so as to detect the illegal building in the video, thereby realizing rapid and accurate identification of the illegal building.

Description

technical field [0001] The invention belongs to the technical field of image recognition target detection, and in particular relates to an accurate detection system for illegal buildings. Background technique [0002] With the continuous development of urbanization, the value of real estate has also risen accordingly. Incidents of illegally occupying land to add illegal buildings or changing the original building design to add illegal buildings are emerging one after another. Due to the concealment of building roofs, illegal buildings are often difficult to find. These illegal buildings have always been a difficult point in the demolition work, but due to limited human resources and information channels, many illegal buildings cannot be quickly discovered and demolished. [0003] At present, with the excellent performance of deep learning in the field of target detection and the popularity of drones, it provides a very powerful solution to the discovery and teardown of illeg...

AI Technical Summary

Problems solved by technology

[0006] After the difference is identified by the method based on image comparison, human judgment may be required. Some current image recognition methods are too inefficient and not suitable for video recognition.

Images