Floor height detection method and related surveillance camera
A floor height detection method is applied to a surveillance camera, and a surveillance image captured by the surveillance camera contains a surveillance region. The floor height detection method includes dividing the surveillance image at least into a first region and a second region, acquiring an actual installing height matrix of the surveillance camera upon the first region and the second region, and utilizing a matrix factorization algorithm to analyze the actual installing height matrix to acquire floor heights within the surveillance region.
1. A floor height detection method applied to a surveillance camera, a surveillance image captured by the surveillance camera containing a surveillance region, the floor height detection method comprising:
dividing the surveillance image at least into a first region and a second region;
acquiring an installing height matrix of the surveillance camera upon the first region and the second region; and
utilizing a matrix factorization algorithm to analyze the installing height matrix for acquiring floor heights of the surveillance region, so as to determine whether to divide the first region or the second region in accordance with the floor heights.
2. The floor height detection method of claim 1 , wherein acquiring the installing height matrix of the surveillance camera upon the first region and the second region comprises:
utilizing the surveillance camera to track and acquire a first calibration object height of a calibration object inside the surveillance region within the first region and further acquire a second calibration object height of the calibration object inside the surveillance region within the second region;
utilizing the first calibration object height and the second calibration object height to acquire a first installation height of the surveillance camera within the first region and further acquire a second installation height of the surveillance camera within the second region; and
generating the installing height matrix via the first installation height and the second installation height.
3. The floor height detection method of claim 2 , wherein the surveillance camera has a predefined installation height, the floor height detection method compares the first calibration object height and the second calibration object height respectively with a reference object height, and computes the first installation height and the second installation height in accordance with a comparison result.
4. The floor height detection method of claim 3 , wherein the floor height detection method acquires the reference object height via an object height statistics result of an area where inside the surveillance camera is located, or via an automatic learning algorithm.
5. The floor height detection method of claim 2 , wherein the calibration object has a known actual height, the floor height detection method analyzes a difference between the first calibration object height and the known actual height to compute the first installation height, and further analyzes a difference between the second calibration object height and the first calibration object height or a difference between the second calibration object height and the known actual height, so as to compute the second installation height.
6. The floor height detection method of claim 1 , wherein at least one calibration object is located inside the surveillance region, the floor height detection method tracks and acquires first variation of a plurality of first calibration object heights of the at least one calibration object within the first region, and compares the first variation with a predefined threshold to determine the floor heights of the first region.
7. The floor height detection method of claim 6 , wherein when the first variation is greater than the predefined threshold, the floor height detection method divides the first region at least into a first sub-region and a second sub-region, and acquires another installing height matrix of the surveillance camera upon the first sub-region and the second sub-region, so as to acquire the floor heights of the first region.
8. The floor height detection method of claim 6 , wherein when the first variation is greater than the predefined threshold, the floor height detection method detects a line-type boundary of the first region, and analyzes change of the plurality of first calibration object heights relative to the line-type boundary, so as to acquire the floor heights of the first region.
9. The floor height detection method of claim 6 , wherein the floor height detection method further tracks and acquires a plurality of second calibration object heights of the at least one calibration object within the second region, and determines the floor heights between the first region and the second region in accordance with analysis of the plurality of second calibration object heights.
10. A surveillance camera, comprising:
an image receiver adapted to receive a surveillance image containing a surveillance region; and
an operation processor electrically connected to the image receiver in a wire manner or in a wireless manner, and adapted to analyze the surveillance image, so as to divide the surveillance image at least into a first region and a second region, acquire an installing height matrix of the surveillance camera upon the first region and the second region, and utilize a matrix factorization algorithm to analyze the installing height matrix for acquiring floor heights of the surveillance region, so as to determine whether to divide the first region or the second region in accordance with the floor heights.
11. The surveillance camera of claim 10 , wherein the operation processor is further adapted to utilize the surveillance camera to track and acquire a first calibration object height of a calibration object inside the surveillance region within the first region and further acquire a second calibration object height of the calibration object inside the surveillance region within the second region, utilize the first calibration object height and the second calibration object height to acquire a first installation height of the surveillance camera within the first region and further acquire a second installation height of the surveillance camera within the second region, and generate the installing height matrix via the first installation height and the second installation height.
12. The surveillance camera of claim 11 , wherein the surveillance camera has a predefined installation height, the operation processor is further adapted to compare the first calibration object height and the second calibration object height respectively with a reference object height, and compute the first installation height and the second installation height in accordance with a comparison result.
13. The surveillance camera of claim 12 , wherein the operation processor is further adapted to acquire the reference object height via an object height statistics result of an area where inside the surveillance camera is located, or via an automatic learning algorithm.
14. The surveillance camera of claim 11 , wherein the calibration object has a known actual height, the operation processor is further adapted to analyze a difference between the first calibration object height and the known actual height to compute the first installation height, and further analyze a difference between the second calibration object height and the first calibration object height or a difference between the second calibration object height and the known actual height, so as to compute the second installation height.
15. The surveillance camera of claim 10 , wherein at least one calibration object is located inside the surveillance region, the operation processor is further adapted to track and acquire first variation of a plurality of first calibration object heights of the at least one calibration object within the first region, and compare the first variation with a predefined threshold to determine the floor heights of the first region.
16. The surveillance camera of claim 15 , wherein when the first variation is greater than the predefined threshold, the operation processor is further adapted to divide the first region at least into a first sub-region and a second sub-region, and acquire another installing height matrix of the surveillance camera upon the first sub-region and the second sub-region, so as to acquire the floor heights of the first region.
17. The surveillance camera of claim 15 , wherein when the first variation is greater than the predefined threshold, the operation processor is further adapted to detect a line-type boundary of the first region, and analyzes change of the plurality of first calibration object heights relative to the line-type boundary, so as to acquire the floor heights of the first region.
18. The surveillance camera of claim 15 , wherein the operation processor is further adapted to track and acquire a plurality of second calibration object heights of the at least one calibration object within the second region, and determine the floor heights between the first region and the second region in accordance with analysis of the plurality of second calibration object heights.