Fried food detection system based on symbiotic double-flow convolutional network and digital image

Abstract

The invention discloses a fried food detection system based on a symbiotic double-flow convolutional network and a digital image. The fried food detection system comprises an image preprocessing module, a rapid identification module, a classification and positioning module, a target cutting module and an image analysis module which are connected in sequence, the image preprocessing module sequentially performs image stylized migration and image filtering processing on the input images to obtain available image tensors of the network; images are rapidly classified through a full convolutional network composed of a symbiotic feature extraction network and an identification network in the rapid identification module; the classification and positioning module is a full convolutional network; the target cutting module cuts a target image from the original image by using the optimal box; and the image analysis module analyzes the target image to give a quantitative analysis result. Accordingto the method, the symbiotic double-flow convolutional network and the digital image analysis are combined, so that quick and accurate fried food positioning and attribute identification can be realized.

Description

technical field [0001] The invention belongs to the field of computer vision and image processing food detection, in particular to a fried food detection system combined with symbiotic double-stream convolution network and digital image analysis technology. Background technique [0002] Fried food is food that people eat daily, such as fried French fries, potato chips, chicken legs, chicken wings, etc. Whether it is a small or medium-sized restaurant, or a large-scale food production workshop or canteen, the fried food is mainly sorted and packaged manually at present; in a high-temperature environment, it is necessary to ensure food hygiene and fast sorting and packaging. Food production enterprises facing increasing labor costs have brought great challenges. Therefore, the study of automatic identification of such fried foods is of great significance to realize the automation of fried foods sorting work in complex environments. [0003] For general sorting work automation...

AI Technical Summary

Problems solved by technology

[0008] (3) As we all know, the training of neural networks requires a large amount of data, and the application scenarios of some detection systems cannot provide sufficient data information such as the detection problem in the food field involved in the present invention

Images