Image detection auxiliary system and image detection method
Disclosed are an image detection auxiliary system and an image detection method. An image capture device is used to generate an original image. A computing device receives the original image and preprocesses the original image as an input image. The computing device uses a classification model to obtain a feature parameter corresponding to the input image, and inputs the feature parameter to a probability function to calculate an image focus score. Finally, the computing device performs formal detection or subsequent processing of generating a second warning signal according to the image focus score.
1 . An image detection auxiliary system, comprising:
an image capture device, capturing at least one object to be detected to generate an original image, wherein before generating the original image, the image capture device further performs an object detection procedure, wherein the object detection procedure comprises:
detecting whether a target object exists in a picture of the image capture device;
in a case that the target object exists in the picture of the image capture device, the image capture device generates the original image; and
in a case that the target object does not exist in the picture of the image capture device, the image capture device generates a first warning signal to display a warning for selecting whether to filter out the picture; and
a computing device, in signal connection to the image capture device to receive the original image, and preprocessing the original image as an input image, where the computing device uses a classification model of the computing device that is trained with a large quantity of images to obtain a feature parameter corresponding to the input image and inputs the feature parameter to a softmax probability function to calculate an image focus score, and the computing device performs subsequent processing according to the image focus score, wherein in a case that the image focus score is lower than a preset standard value, the computing device generates a second warning signal to warn that the image focus score does not meet a standard, wherein the subsequent processing includes adjusting a focus parameter of the image capture device.
2 . The image detection auxiliary system according to claim 1 , further comprising a display device, in signal connection to the computing device, and configured to display the input image and the image focus score.
3 . The image detection auxiliary system according to claim 1 , wherein the step of the computing device preprocessing the original image further comprises:
cropping a target region from the original image; and
adjusting a size of the target region as the input image.
4 . The image detection auxiliary system according to claim 3 , wherein the target region is a central region of the original image.
5 . The image detection auxiliary system according to claim 1 , wherein the image capture device detects whether the target object exists in the picture through an object detection model.
6 . The image detection auxiliary system according to claim 1 , wherein in a case that the image focus score reaches a preset standard value, the computing device performs formal detection on the input image.
7 . The image detection auxiliary system according to claim 6 , wherein the formal detection comprises surface defect detection, image classification processing, or image segmentation processing.
8 . An image detection method, comprising:
capturing at least one object to be detected to generate an original image by an image capture device, wherein before the step of generating the original image, the method further comprises:
performing an object detection procedure, wherein the object detection procedure comprises:
detecting whether a target object exists in a picture of the image capture device;
in a case that the target object exists in the picture of the image capture device, the image capture device generates the original image; and
in a case that the target object does not exist in the picture of the image capture device, the image capture device generates a first warning signal to display a warning for selecting whether to filter out the picture;
preprocessing the original image as an input image by a computing device;
using a classification model of the computing device that is trained with a large quantity of images to obtain a feature parameter corresponding to the input image;
inputting the feature parameter to a softmax probability function to calculate an image focus score; and
performing subsequent processing according to the image focus score, wherein in the step of performing the subsequent processing according to the image focus score, in a case that the image focus score is lower than a preset standard value, generating a second warning signal to warn that the image focus score does not meet a standard, wherein the subsequent processing includes adjusting a focus parameter of the image capture device.
9 . The image detection method according to claim 8 , wherein the step of preprocessing the original image further comprises:
cropping a target region from the original image; and
adjusting a size of the target region as the input image.
10 . The image detection method according to claim 9 , wherein the target region is a central region of the original image.
11 . The image detection method according to claim 8 , wherein the image capture device detects whether the target object exists in the picture through an object detection model.
12 . The image detection method according to claim 8 , wherein in the step of performing the subsequent processing according to the image focus score, in a case that the image focus score reaches a preset standard value, the subsequent processing comprises: performing formal detection on the input image.
13 . The image detection method according to claim 12 , wherein the formal detection comprises surface defect detection, image classification processing, or image segmentation processing.