Target object detection method and device

Abstract

The invention provides a target object detection method, which comprises the steps of respectively obtaining a visible light image and an infrared image through the camera equipment of a binocular stereo vision system; according to the pre-calibrated parameters of the camera equipment, detecting the image of a target object in the visible light image and in the infrared image through a preset detection algorithm; and obtaining the depth information of the target object according to the detected position of the target object and the spatial geometry information of the image of the target object. The invention also provides a target object detection device, wherein based on the fusion of images obtained by different data sources, the target object is detected. Meanwhile, the depth information of the target object is obtained. Therefore, the detection recall rate and the detection accuracy rate of the target object are improved.

Description

technical field [0001] The invention relates to the technical field of computer vision, in particular to a method and device for detecting a target object. Background technique [0002] Due to its key role in the semi-automatic and automatic driving fields of automobiles, the automatic detection of target objects has been a hot research topic in the field of computer vision (CV, ComputerVision) for many years. Computer vision is a science that studies how to make machines "see". It refers to machine vision that uses cameras and computers instead of human eyes to identify, track and measure target objects, and further performs graphics processing to make computer processing more suitable for human eyes. Images that are observed or transmitted to the instrument for detection. [0003] In the prior art, people use various methods, such as infrared, radar, visible light, etc., to obtain the position of the detection target object and the depth of the target object. However, th...

AI Technical Summary

Problems solved by technology

However, the data sources involved in the above detection methods have their own disadvantages.

Specifically, the data obtained by infrared means is suitable for night work, but the working distance is limited, and it can only work at a relatively short distance, and the application range is narrow

The data obtained by radar is less reliable when the target object is far away, and it is easily affected by the weather, such as being interfered by raindrops, and the effect is poor

The image data obtained by infrared and visible light is ubiquitous. When using simple algorithms to process the images, the effect is not ideal, and the algorithm complexity to achieve the ideal effect is high, and it is difficult to apply in practice.

Images