Image Recognition Method for Product Quantity Detection

Abstract

The invention discloses an image identification method for detecting product quantity, and relates to an image identification method. The image identification method comprises the following steps of: S1, installing video equipment; S2, packaging products; S3, stacking packaging boxes; S4, acquiring videos; S5, performing video screenshot; S6, performing preprocessing on pictures; S7, performing intermediate identification on the preprocessed pictures; and S8, judging the total quantity of the products. The image identification method is simple and high in identification accuracy, the production cost can be reduced, labor intensity can be reduced, the production efficiency can be improved, and the image identification method can be popularized and applied to any just-in-time production enterprises.

Description

technical field [0001] The invention relates to an image recognition method, in particular to an image recognition method for product quantity detection. Background technique [0002] With the development of computer technology and information technology, image recognition technology has been actively promoted and applied, such as the analysis and recognition of various medical pictures in medical diagnosis, satellite cloud image recognition in weather forecast, fingerprint recognition, face recognition, etc., image recognition Technology is increasingly permeating our everyday lives. However, most of the existing image recognition methods have the following disadvantages: there are many steps in the whole process of image recognition, each step has several algorithms, and the quality of the algorithm directly affects the effect of the subsequent steps, and the processing of images with poor quality is not easy. Ideally, the algorithm is less efficient. For example, in ima...

AI Technical Summary

Problems solved by technology

However, most of the existing image recognition methods have the following disadvantages: there are many steps in the whole process of image recognition, each step has several algorithms, and the quality of the algorithm directly affects the effect of the subsequent steps, and the processing of images with poor quality is not easy. Ideal, the efficiency of the algorithm is low

For example, in image segmentation, it is difficult to achieve the ideal segmentation effect by using a single feature image segmentation method. If multiple image segmentation methods are used in combination, it will often cause excessive segmentation of the image, which is not conducive to image recognition; some algorithms use Gabor Transformation is used to enhance the image, but due to the insensitivity to the feature point area, the enhancement effect of this part is not good, and it is often impossible to distinguish the end point from the branch point, which makes this feature information invalid

[0003] In addition, in industrial production, the detection of product quantity by enterprises is generally carried out by manual intervention. This method has the following disadvantages: more labor needs to be invested, higher production costs, higher labor intensity, inventory accuracy and production less efficient

However, if the traditional image recognition technology is used for the counting and detection of enterprise product quantities, there will be shortcomings such as low efficiency, low accuracy, and overly complicated methods.

Images