Method of detecting cloud deck camera working position by using motion vector method

Abstract

The invention relates to a method of detecting cloud deck camera working position by using a motion vector method. The method comprises the following steps of: firstly, sending one control instruction to a cradle head camera, cotinuously sampling and preserving the sent control instruction, and continuing N frame image sequence for one period; secondly, calculating a gradient map of the current frame, carrying out binarization on the gradient map, regarding a pixel the gradient of which is less than a threshold value as a pixel of a flat site, forming a two-value mask of the flat site, and carrying out morphology filtering on the flat site mask; calculating a motion vector of the current frame on a block of the motion vector to be calculated in the current frame according to the preview frame; and finally aiming at different cradle head control instructions, calculating the motion vector characteristic of continuous N-1 frames, and determining a working position of the cradle head camera. In the method provided by the invention, the motion vector of the image sequence is adopted to estimate and detect whether the motor function of the cradle head camera is normal. The method only needs the existing cradle head camera, does not needs an additional hardware unit, and can automatically detect the working condition of the cradle head camera.

Description

technical field [0001] The invention relates to the technical field of automatic detection, in particular to a method for automatic fault detection of a pan-tilt camera for video monitoring. Background technique [0002] PTZ cameras are widely used in video surveillance, industrial monitoring, industrial robots and other fields. The number and running time of PTZ cameras are increasing day by day, and the workload of equipment maintenance and fault detection has become increasingly complex and huge with the expansion of the system. The traditional manual inspection method is inefficient, and an automatic fault detection method for PTZ cameras is urgently needed in applications. method of detection. [0003] The existing method "Pan-Tilt self-diagnosis and monitoring circuit (Invention Publication No. 101776914A)" requires an additional hardware circuit to realize the function of automatic detection of PTZ faults, which increases the system cost. [0004] "Pan-tilt detectio...

AI Technical Summary

Problems solved by technology

The number and running time of PTZ cameras are increasing day by day, and the workload of equipment maintenance and fault detection has become increasingly complex and huge with the expansion of the system. The traditional manual inspection method is inefficient, and an automatic fault detection method for PTZ cameras is urgently needed in applications. detection method

[0003] The existing method "Pantai self-diagnosis monitoring circuit (Invention Publication No. 101776914A)" requires additional hardware circuits to realize the function of automatic detection of PTZ faults, and the system cost increases

[0004] "Pan-tilt detection method and detection device (101631260)", "Image comparison method for detecting the working status of the pan-tilt camera (101420593)", "Automatic camera fault detection and management method, device and system (101783970A)" using the pan-tilt Whether the difference between the two frames of image data before and after the control command is sent exceeds the threshold, and judge whether the movement of the pan / tilt is normal; this type of method is greatly affected by image noise and interference, and may easily cause false detection

[0005] "A method and device for automatic monitoring of front-end pan-tilt failure of video surveillance equipment (101098466)" judges whether the camera is moving by comparing the difference between the current image and the static area of ​​the static background image; this method is easily affected by local moving objects, resulting in False detection

[0006] "A video quality diagnosis system and its implementation method (102176758A)" mentioned the diagnosis of PTZ control, in which the image perspective transformation matching technology is used to detect the movement of the picture; this method is easily affected by factors such as noise interference and object movement, and is reliable Sex is not high

Images