Evaluation apparatus, evaluation method, and evaluation program
An evaluation apparatus, an evaluation method, and an evaluation program applicable to a quality evaluation of sintered bodies are provided. The evaluation apparatus includes: an acquisition unit configured to acquire a cross-sectional image obtained by photographing a cross-section of a sintered body group stained with a staining solution; a generation unit configured to extract a saturation component for each of sintered bodies from the cross-sectional image, thereby to generate a saturation component image; and a visualization unit configured to visualize the saturation component image.
1 . An evaluation apparatus comprising:
a memory; and
a processor coupled to the memory and configured to:
acquire a cross-sectional image obtained by photographing a cross-section of a sintered body group stained with a staining solution;
extract a saturation component for each of sintered bodies from the cross-sectional image, thereby to generate a saturation component image; and
visualize the saturation component image.
2 . The evaluation apparatus according to claim 1 , wherein the processor is configured to convert each pixel value of a separate image that is extracted for each of the sintered bodies from the cross-sectional image, from an (Red, Green, Blue) (RGB) color space to an (Hue, Saturation, Value) (HSV) color space, and to extract a saturation component from the converted each pixel value, thereby to generate the saturation component image.
3 . The evaluation apparatus according to claim 2 , wherein
the cross-sectional image is obtained by cutting a resin in which the sintered body group is embedded, staining the cross-section obtained by the cutting with the staining solution, and photographing the cross-section, and
the processor is configured to extract a region of the sintered bodies exposed in the cross-section as a foreground region from the cross-sectional image, thereby to extract the separate image for each of the sintered bodies from the cross-sectional image.
4 . The evaluation apparatus according to claim 3 , wherein the processor is configured to process a background region other than the foreground region in the cross-sectional image, and to extract the foreground region from the processed cross-sectional image.
5 . The evaluation apparatus according to claim 1 , wherein the processor is configured to generate a two-dimensional image in which a value of a saturation component of each of pixels of the saturation component image is assigned to a specific color, thereby to visualize the saturation component image.
6 . The evaluation apparatus according to claim 5 , wherein the processor is configured to generate a three-dimensional image in which the value of the saturation component of each of the pixels of the saturation component image is assigned to the specific color and also assigned to a coordinate in a height direction in a three-dimensional space, thereby to visualize the saturation component image.
7 . The evaluation apparatus according to claim 6 , further comprising a storage in which an inspection result for each of the sintered bodies, which is input in response to displaying a display screen including one or both of the two-dimensional image and the three-dimensional image, is stored in association with the saturation component image.
8 . The evaluation apparatus according to claim 7 , wherein the inspection result is any one of: a crack of a first depth level is included in a sintered body; a crack of a second depth level shallower than the first depth level is included in the sintered body; or no cracks are included in the sintered body.
9 . The evaluation apparatus according to claim 8 , wherein evaluation result information including an aggregate result and a calculation result is output, in which the aggregate result is obtained by aggregating, for the sintered body group, the inspection result for each of the sintered bodies, and in which the calculation result is obtained by calculating a percentage of sintered bodies that include the crack of the first depth level.
10 . The evaluation apparatus according to claim 8 , wherein a training process is performed on a training model using training data read from the storage with the saturation component image as input data and the inspection result in association with the saturation component image as ground-truth data.
11 . The evaluation apparatus according to claim 8 , wherein a trained model, which has been trained by a training process using training data read from the storage with the saturation component image as the input data and the inspection result in association with the saturation component image as the ground-truth data, receives as input a saturation component image generated from a newly photographed cross-sectional image, to predict an inspection result.
12 . The evaluation apparatus according to claim 11 , wherein evaluation result information including an aggregate result and a calculation result is output, in which the aggregate result is obtained by aggregating, for the sintered body group, the inspection result predicted for each of the sintered bodies, and in which the calculation result is obtained by calculating a percentage of sintered bodies that include the crack of the first depth level.
13 . An evaluation method comprising:
acquiring a cross-sectional image obtained by photographing a cross-section of a sintered body group stained with a staining solution;
extracting a saturation component for each of sintered bodies from the cross-sectional image, thereby generating a saturation component image; and
visualizing the saturation component image.
14 . A non-transitory computer-readable recording medium storing a program for causing a computer to execute:
acquiring a cross-sectional image obtained by photographing a cross-section of a sintered body group stained with a staining solution;
extracting a saturation component for each of sintered bodies from the cross-sectional image, thereby generating a saturation component image; and
visualizing the saturation component image.