Object size measurement method and image database data acquisition method

Abstract

The invention discloses an object size measurement method. A calibration object is placed near a target object, pixel sizes of the target object and the calibration object are calculated, the actual size of the target object is calculated, no measurement tool is needed, and the external size of a general object can be quickly and accurately measured. Meanwhile, the invention also discloses an image database data acquisition method based on the object size measurement method. When a database comprising a target object image sample and the size of the target object is built, automatic database construction can be realized, that is, the target object image sample and the size can be obtained at the same time at one time through photographing, corresponding database construction is realized, and compared with manual recording, the method of the invention is labor-saving and time-saving, and is less error-prone.

Description

technical field [0001] The invention relates to the field of data collection, in particular to a method for measuring physical dimensions and a data collection method for an image database. Background technique [0002] In the field of content-based image recognition and augmented reality technology, it is usually necessary to collect image data in advance to establish an image database. The conventional method is to take pictures of the objects to be collected one by one and store them in the database. If it is necessary to add other parameters of the object to the database, such as size, Size, etc., usually need to be manually measured, and then manually matched to the photo input, manual measurement and input work is very heavy, and manual error-prone. [0003] In terms of object size measurement, there are studies on the size of large objects in the prior art, for example: "Shi Mengmeng, Han Yan, Wang Jian, etc. Research on 3D Measurement Method of Large Object Size Base...

AI Technical Summary

Problems solved by technology

[0002] In the field of content-based image recognition and augmented reality technology, it is usually necessary to collect image data in advance to establish an image database. The conventional method is to take pictures of the objects to be collected one by one and store them in the database. If it is necessary to add other parameters of the object to the database, such as size, Size, etc., usually need to be manually measured, and then manually matched to the photo input, manual measurement and input work is very heavy, and manual error-prone

[0003] In terms of object size measurement, there are studies on the size of large objects in the prior art, for example: "Shi Mengmeng, Han Yan, Wang Jian, etc. Research on 3D Measurement Method of Large Object Size Based on Image, Computer Measurement and Control, 2011019( 10): 2351-2353", when measuring the size of a large object mentioned in this article, it is necessary to adaptively find a calibration template from the object itself, and then manually measure the size of the calibration template. Every time an object is measured, it is necessary to manually measure the calibration template once , this method has no problem in measuring large objects, but if it is applied to the measurement of a large number of ordinary objects, it is more complicated than manual direct measurement, and it is not suitable for objects with no significant features on the surface of the target measurement object. Possibility to calibrate the template, so this method has limited scope

Images