Abstract: A determination unit determines, on images of video data where a road set as a survey target was shot in different times, whether or not the shooting position of the image is within a tolerance value with the shooting position of any of the images as a reference, for each image having a corresponding shooting position. When the shooting position is determined to be within the tolerance value, a creation unit creates screen information of a screen where images that has been determined to be within the tolerance value is displayed in synchronization. Moreover, when the shooting position is determined to be beyond the tolerance value, the creation unit creates screen information of a screen where an image that has been determined to be beyond the tolerance value is undisplayed.

Abstract: An image processing apparatus includes a processor that executes a process. The process includes converting an input image into multiple types of images based on values of pixels in the input image, and selecting at least one image from the multiple types of images based on luminance information of the multiple types of images.

Abstract: A road surface inspection device detects an abnormal region on a road surface from the road surface on a road image captured by a camera. The road surface inspection device determines whether the detected abnormal region overlaps a trace on the road image, along which the wheels of a patrol car are expected to pass on the road surface. When the abnormal region overlaps the trace of wheels, the road surface inspection device determines whether an acceleration at the measurement time corresponding to the captured time of the road image among the accelerations measured during driving on the road surface by a G-sensor is outside a predetermined range.

Abstract: A color classification method including: determining a plurality of predetermined process target regions from a plurality of pickup images taken by an imaging device; calculating a plurality of color distributions of pixels contained in the respective determined process target regions; and forming a plurality of clusters by executing a clustering process based on the calculated color distributions for the pickup images, the method being for classifying a plurality of colors with respect to the respective formed clusters, the forming step including: extracting predetermined number of the classified colors, whose calculated rates are highest in the calculated plurality of the rates, from among the predetermined classified colors; defining a plurality of color spaces, dimensions of each of which are selected from the predetermined number of the extracted classified colors.

Abstract: This invention is a technique to presume colors of an image. A color extraction method according to this invention include a luminance calculation step of calculating a luminance in a color extraction region, which is included in an image and from which a color is to be extracted, and storing the calculated luminance into a storage device; a color name extraction of referring to a luminance database storing a color name and a luminance corresponding to the color name to extract a color name in the color extraction region by using, as a main judgment element, the luminance in the color extraction region, which is stored in the storage device. As a result, it is possible to presume a color even in a case of the monochromatic image or color image photographed in the night.

Abstract: A color classification method including: determining a plurality of predetermined process target regions from a plurality of pickup images taken by an imaging device; calculating a plurality of color distributions of pixels contained in the respective determined process target regions; and forming a plurality of clusters by executing a clustering process based on the calculated color distributions for the pickup images, the method being for classifying a plurality of colors with respect to the respective formed clusters, the forming step including: extracting predetermined number of the classified colors, whose calculated rates are highest in the calculated plurality of the rates, from among the predetermined classified colors; defining a plurality of color spaces, dimensions of each of which are selected from the predetermined number of the extracted classified colors.

Abstract: This invention is a technique to presume colors of an image. A color extraction method according to this invention include a luminance calculation step of calculating a luminance in a color extraction region, which is included in an image and from which a color is to be extracted, and storing the calculated luminance into a storage device; a color name extraction of referring to a luminance database storing a color name and a luminance corresponding to the color name to extract a color name in the color extraction region by using, as a main judgment element, the luminance in the color extraction region, which is stored in the storage device. As a result, it is possible to presume a color even in a case of the monochromatic image or color image photographed in the night.