System and method for obstacle detection using fusion of color space information

A system comprising:

• an image sensor to obtain an image of an area, the obtained image representing the area in a first color space;
• and
• a processing unit to convert the obtained image from the image sensor to at least one second color space;
• wherein converting the obtained image to at least one second color space produces a plurality of converted images, each of the plurality of converted images corresponding to one of a plurality of color sub-spaces in the at least one second color space;
• wherein the processing unit calculates upper and lower thresholds for at least two of the plurality of color sub-spaces, segments the converted images corresponding to the at least two color sub-spaces based on the corresponding upper and lower thresholds, segments the image obtained by the image sensor based on the at least two segmented converted images, and filters the segmented image obtained from the image sensor based on edge density data in the image obtained from the image sensor;
• wherein the processing unit determines the edge density data based on edges identified in respective regions of the received image, each region enclosing a respective detected obstacle.

A program product comprising a non-transitory processor-readable medium on which program instructions are embodied, wherein the program instructions are operable, when executed by at least one programmable processor, to cause the programmable processor to:

• convert a received image of an area from a first color space to at least one second color space to produce a plurality of converted images, each of the plurality of converted images corresponding to one of a plurality of color sub-spaces of the at least one second color space;
• calculate an upper threshold and a lower threshold for each of at least two of the plurality of color sub-spaces;
• apply the upper and lower thresholds to the converted images corresponding to the at least two color sub-spaces to segment the corresponding converted images;
• fuse the segmented converted images corresponding to the at least two color sub-spaces to segment the received image;
• and
• update the segmentation of the received image based on edge density data in the received image;
• wherein the edge density data is determined based on edges identified in respective regions of the received image, each region enclosing a respective detected obstacle.