Extraction of image feature data from images

1. A method for obtaining image features of an image, the method comprising:

extracting from the image, by a processor, a color histogram of the image, the extraction of the color histogram including performing one-dimensional sampling of pixels comprising the image in each of a first dimension of a color space, a second dimension of the color space, and a third dimension of the color space;

analyzing an edge map corresponding to the image to detect a pattern included in the image;

in response to a confidence level of the pattern detection being below a pre-defined threshold, extracting from the image an orientation histogram of the image; and

extracting from the image a dominant color of the image.

2. The method of claim 1 , wherein the color space comprises a hue, saturation, and value (HSV) color space and a gray channel, and wherein the first dimension comprises hue, the second dimension comprises saturation, and the third dimension comprises value, and a fourth dimension comprises the gray channel.

3. The method of claim 2 , wherein the performing of the one-dimensional sampling comprises sampling the pixels in the first dimension in 24 bins, sampling the pixels in the second dimension in 8 bins, sampling the pixels in the third dimension in 8 bins, and sampling the pixels in the fourth dimension in 8 bins.

4. The method of claim 2 , wherein the extraction of the color histogram comprises applying a weight to each of a one-dimensional sample in the first dimension, a one-dimensional sample in the second dimension, a first portion of a one-dimensional sample in the third dimension having a saturation level above a pre-defined threshold, and a second portion of the one-dimensional sample in the third dimension having the saturation level below the pre-defined threshold, wherein the second portion of the one-dimensional sample in the third dimension corresponds to the gray channel.

5. The method of claim 4 , wherein the extraction of the color histogram comprises stacking the weighted one-dimensional samples to form the color histogram.

6. The method of claim 1 , wherein the performing of the one-dimensional sampling comprises performing uniform or non-uniform one-dimensional sampling.

7. The method of claim 1 , wherein the dominant color is extracted in response to the orientation histogram being indicative of the image having low spatial variation.

8. The method of claim 1 , wherein the edge map comprises a Canny edge map.

9. The method of claim 1 , further comprising, in response to the confidence level of the pattern detection being at or above the pre-fined threshold, indexing the image in accordance with at least one of the confidence lever, the color histogram, or the pattern.

10. The method of claim 1 , further comprising determining a confidence score for the dominant color based on the color histogram, and indexing the image in accordance with at least one of the confidence score, the color histogram, or the dominant color.

11. The method of claim 1 , further comprising, in response to the orientation histogram being indicative of the image having high spatial variation, determining a confidence score for an orientation feature of the image and indexing the image in accordance with at least one of the confidence score, the color histogram, the orientation histogram, or the orientation feature.

12. The method of claim 1 , wherein the image comprises an inventory image corresponding to an inventory item offered for sale by an electronic commerce (e-commerce) site or online marketplace, a photograph of an item of interest submitted by a user, or an image included in a web page that is of interest to the user.

13. The method of claim 1 , further comprising identifying a sampling area of the image, the sampling area comprising a central portion of the image to be used for extraction of the color histogram, the pattern detection, extraction of the orientation histogram, and extraction of the dominant color.

14. A device, comprising:

at least one memory;

at least one processor in communication with the memory; and

one or more modules comprising instructions stored in the memory and executed by the processor to perform operations comprising:

extracting from an image a color histogram of the image, the extraction of the color histogram including performing one-dimensional sampling of pixels comprising the image in each of a first dimension of a color space, a second dimension of the color space, and a third dimension of the color space;

analyzing an edge map corresponding to the image to detect a pattern included in the image;

in response to a confidence level of the pattern detection being below a pre-defined threshold, extracting from the image an orientation histogram of the image; and

extracting from the image a dominant color of the image.

15. The device of claim 14 , wherein the device comprises a server in communication with a mobile device configured to provide user interface screens to receive inputs from a user.

16. The device of claim 15 , wherein the image comprises a photograph taken using the mobile device and received by the device from the mobile device.

17. The device of claim 14 , wherein the image comprises an inventory image corresponding to an inventory item offered for sale by an electronic commerce (e-commerce) or online marketplace.

18. The device of claim 14 , wherein the color space comprises a hue, saturation, and value (HSV) color space and a gray channel, and wherein the performing of the one-dimensional sampling comprises performing uniform one-dimensional sampling.

19. A non-transitory computer readable medium including instructions, when ex cured by a processor, causes the processor to perform operations comprising:

extracting from an image a color histogram of the image, the extraction of the color histogram including performing one-dimensional sampling of pixels comprising the image in each of a first dimension of a color space, a second dimension of the color space, and a third dimension of the color space;

analyzing an edge map corresponding to the image to detect a pattern included in the image;

in response to a confidence level of the pattern detection being below a pre-defined threshold, extracting from the image an orientation histogram of the image; and

extracting from the image a dominant color of the image.

20. The non-transitory computer readable medium of claim 19 , wherein extraction of the orientation histogram comprises:

calculating an x-derivative and a v-derivative of the edge map;

calculating a gradient and an orientation using the x-derivative and the y-derivative; and

applying a weight to each edge pixel included in the edge map to obtain the orientation histogram.