Vision based pedestrian and cyclist detection method

1. A vision based pedestrian and cyclist detection method, proceeded by a computing processor, comprising the following steps of:

receiving an input image containing at least one human and at least one two-wheeled vehicle;

calculating a pixel value difference between each pixel and the neighbor pixels thereof within the input image;

quantifying the pixel value difference of each pixel, and defined as a weight of the pixel;

proceeding statistics for the pixel value differences and the weights;

determining intersections of the statistics as a feature of the input image;

classifying the feature into human feature and non-human feature;

proceeding a two-wheeled vehicle detection to the input image;

confirming whether the human feature is belonging to cyclist if when the spatial relationship between the human feature and the detected two-wheeled vehicle holds; and

retaining one detection result for each cyclist by suppressing other weaker spatial relationships between the human feature and the detected two-wheeled vehicle.

2. The vision based pedestrian and cyclist detection method as claimed in claim 1 , wherein the pixel value difference is encoded as an 8-bit binary number.

3. The vision based pedestrian and cyclist detection method as claimed in claim 2 , wherein the weight of the pixel is defined by a following formula:

weight

=

∑

i

=

0

7

⁢



I

y

i

-

I

x



⁢

,

where I is a pixel value, x is the pixel, and y is the neighbor of the pixel.

4. The vision based pedestrian and cyclist detection method as claimed in claim 3 , wherein the result of statistics is a set of histograms, each histogram is constructed according to the pixel value difference and the weight of pixel.

5. The vision based pedestrian and cyclist detection method as claimed in claim 4 , wherein classifying the feature into human feature and non-human feature is achieved by a pre-trained linear supported vector machine (SVM) classifier.

6. The vision based pedestrian and cyclist detection method as claimed in claim 5 , wherein the spatial relationship between the human feature and the detected two-wheeled vehicle is described by a horizontally geometric mean, a vertically geometric mean and a relative ratio, of the human feature and the detected two-wheeled vehicle.

7. The vision based pedestrian and cyclist detection method as claimed in claim 5 , wherein the spatial relationship between the human feature and the detected two-wheeled vehicle is converted into a probability to be compared with a predetermined threshold to confirm whether the human feature belongs to cyclist.