Motion information-based method for monocular camera to acquire depth information

Abstract

The invention discloses a motion information-based method for a monocular camera to acquire the depth information. The method comprises the following steps of taking at least two pictures different in sharpness by means of a camera during the movement process of the camera, wherein the pictures are the same in illumination condition; secondly, conducting the blocking treatment on the obtained pictures and selecting small blocks of images that are required for computation; conducting the filtering treatment on each block of clear pictures according to different fuzzy matrixes, and finding out a fuzzy matrix highest in level of similarity through matching fuzzy images; obtaining the depth information of the block based on the camera imaging principle by utilizing the above fuzzy matrix highest in level of similarity; repeating the above steps to obtain the depth information of the entire scene. By adopting the above method, the front and rear distances of an object in a picture can be calculated based on a simple algorithm. The method is applicable to both the short-distance measurement and the long-distance measurement.

Description

technical field [0001] The invention relates to a method for extracting depth data, in particular to a method for obtaining depth information by a monocular camera based on motion information. Background technique [0002] At present, there are mainly two depth data extraction methods, one is to use a depth camera to capture pictures to obtain the depth data of the picture. Currently, the commonly used implementation methods of depth cameras include TOF (Timeofflight), structured light, and laser scanning. Among them, TOF cameras are used more, which are mainly used in applications such as robots and interactive games. TOF is the 3D imaging of the optical time-of-flight method, which is to continuously send light pulses to the target, and then use the sensor to receive the light returned from the object, and obtain the distance of the target object by detecting the flight (round-trip) time of the light pulse. The absolute accuracy of the TOF camera is + 15mm. However, t...

AI Technical Summary

Problems solved by technology

However, in practical applications, the camera layout space is limited, and the baseline length is only a few meters or tens of meters at most, which leads to low accuracy of binocular ranging under long-distance conditions; and because the image quality is affected by external lighting, etc. The conditions are too restrictive and the performance is not stable enough; in addition, the algorithm is relatively complex and other reasons, which limit its application, and in the binocular distance measurement, it must be ensured that all the obtained images are clear images, and the blurred images cannot be further processed but only used as invalid image drop

Images