Ice cream image training method, detection method and processing equipment
A training method and image detection technology, which is applied in the field of image recognition, can solve problems such as the inability to realize automatic counting and statistics of ice cream, difficulty in accurate recognition, etc., and achieve the effect of realizing automatic statistics and improving accuracy
- Summary
- Abstract
- Description
- Claims
- Application Information
AI Technical Summary
Problems solved by technology
Method used
Image
Examples
Embodiment 1
[0028] Such as figure 1As shown, the present invention provides an ice cream image training method, including the following steps: S100: Obtain multiple ice cream sample images of a real scene and label the ice cream, and obtain multiple ice cream labeled images with original mask labels. Multiple ice cream sample images in real scenes are obtained by shooting freezers with various ice creams. There should be a large visual difference between the ice cream sample images, which can be achieved through differences in ambient light intensity, color, angle, and direction. difference in order to improve the training effect. Labeling the ice cream is to distinguish the ice cream from the shooting background. Manual labeling or automatic labeling methods in the prior art can be used for automatic labeling to improve efficiency. S200: Select the labeled image of ice cream, and extract all ice cream block images according to the original mask label. Annotate the ice cream image, and ...
Embodiment 2
[0036] An ice cream image detection method, the ice cream image is detected by the ice cream image training method in Embodiment 1, and the actual mask label of the ice cream covered in the ice cream image is obtained. The detection method can automatically identify different parts of the same ice cream image after being occluded as the same ice cream, and segment the positions of different parts, thereby improving the accuracy of ice cream recognition.
[0037] As an optional implementation, the detection method can also mark the unoccluded surface ice cream in the ice cream image to obtain the surface mask label, which corresponds to the complete ice cream. Labeling the ice cream is to distinguish the ice cream in the ice cream image from the shooting background, and manual labeling or automatic labeling methods in the prior art can be used for automatic labeling to improve efficiency.
[0038] As an optional embodiment, the ice cream in the ice cream image can be counted by...
Embodiment 3
[0040] The present invention also provides an embodiment of a processing device, including one or more processors and memory; wherein, the memory is used to store one or more computer programs, and the one or more processors are used to execute one or more computer programs stored in the memory. program, so that the processor executes the method for detecting ice cream in the second embodiment above. The processing equipment effectively solves the problem of segmenting and identifying different parts of the same ice cream into multiple ice creams due to mutual occlusion between ice creams in the ice cream image, improves the accuracy of ice cream image segmentation and identification in the freezer, and realizes automatic control of the number of ice creams. statistics.
PUM
Login to View More Abstract
Description
Claims
Application Information
Login to View More 
