Image processing apparatus, image processing method, and computer-readable recording device

1. An image processing apparatus for performing positioning between a plurality of spectral band images obtained by capturing images inside a lumen by using a plurality of rays of light having wavelength bands different from one another, the image processing apparatus comprising:

a processor comprising hardware, wherein the processor is configured to:

acquire the plurality of spectral band images;

extract feature data for each spatial frequency band from each pixel in at least one spectral band image of the plurality of spectral band images;

extract, for each spatial frequency band, a pixel area whose spatial frequency component is negative, from the at least one spectral band image, based on the feature data for each spatial frequency band extracted from each pixel in the at least one spectral band image;

calculate, for each spatial frequency band, weights for enhancing a spatial frequency component in a spatial frequency band including information in which a pixel value profile is concave as compared with surrounding areas; and

perform positioning between the plurality of spectral band images based on the weights for each spatial frequency band.

2. The image processing apparatus according to claim 1 ,

wherein the processor is configured to:

extract a plurality of local areas from the at least one spectral band image;

calculate local frequency feature data which is feature data for each spatial frequency band in each of the plurality of local areas, from the feature data for each spatial frequency band extracted from each pixel in the at least one spectral band image; and

calculate weights for each spatial frequency band given to each of the plurality of local areas, as local weights, based on the local frequency feature data.

3. The image processing apparatus according to claim 2 ,

wherein the processor is configured to calculate the local weights based on a wavelength band of the at least one spectral band image in addition to the local frequency feature data.

4. The image processing apparatus according to claim 3 ,

wherein the processor is configured to enhance or suppress a local weight in a spatial frequency band selected according to the wavelength band from among the local weights calculated based on the local frequency feature data.

5. The image processing apparatus according to claim 4 ,

wherein the processor is configured to:

perform threshold processing on the local frequency feature data of the spatial frequency band selected according to the wavelength band; and

multiply the weight calculated based on the local frequency feature data by a specified coefficient, in the spatial frequency band where the local frequency feature data is determined to be greater than or equal to a specified threshold value by the processor.

6. The image processing apparatus according to claim 2 ,

wherein the processor is configured to:

extract a negative area which is a pixel area where amplitude of a spatial frequency component is negative, from the at least one spectral band image;

calculate an average value of absolute values of amplitude in the negative area, for each spatial frequency band, in each of the plurality of local areas; and

determine the average value calculated for each spatial frequency band as the local frequency feature data in each of the plurality of local areas.

7. The image processing apparatus according to claim 2 ,

wherein the processor is configured to:

extract spatial frequency bands whose local frequency feature data is greater than or equal to a specified threshold value;

extract a spatial frequency band whose local frequency feature data is minimum from among the spatial frequency bands extracted by the processor; and

calculate a ratio of the local frequency feature data of the spatial frequency bands extracted by the processor to the local frequency feature data of the spatial frequency band extracted by the processor.

8. The image processing apparatus according to claim 1 ,

wherein the processor is configured to create a template by enhancing a pixel value of each pixel in the at least one spectral band image by using the weights for each spatial frequency band and to perform the positioning between the plurality of spectral band images by using the template.

9. An image processing method for performing positioning between a plurality of spectral band images obtained by capturing images inside a lumen by using a plurality of rays of light having wavelength bands different from one another, the image processing method comprising:

acquiring the plurality of spectral band images;

extracting feature data for each spatial frequency band from each pixel in at least one spectral band image of the plurality of spectral band images;

extracting, for each spatial frequency band, a pixel area whose spatial frequency component is negative, from the at least one spectral band image, based on the feature data for each spatial frequency band extracted from each pixel in the at least one spectral band image;

calculating, for each spatial frequency band, weights for enhancing a spatial frequency component in a spatial frequency band including information in which a pixel value profile is concave as compared with surrounding areas; and

performing positioning between the plurality of spectral band images based on the weights for each spatial frequency band.

10. A computer-readable recording device with an executable program stored thereon, the program instructing a processor for performing positioning between a plurality of spectral band images obtained by capturing images inside a lumen by using a plurality of rays of light having wavelength bands different from one another, to execute:

acquiring the plurality of spectral band images;

extracting feature data for each spatial frequency band from each pixel in at least one spectral band image of the plurality of spectral band images;

extracting, for each spatial frequency band, a pixel area whose spatial frequency component is negative, from the at least one spectral band image, based on the feature data for each spatial frequency band extracted from each pixel in the at least one spectral band image;

calculating, for each spatial frequency band, weights for enhancing a spatial frequency component in a spatial frequency band including information in which a pixel value profile is concave as compared with surrounding areas; and

performing positioning between the plurality of spectral band images based on the weights for each spatial frequency band.