Needs analysis experience - you're right team

Project Objectives

　　With the development of artificial intelligence, supermarkets and retail stores and other retail outlets there have been utilizing AI techniques for real-time monitoring of the shelf, AI can automatically identify the type of each commodity shipments shelves, placement and the remaining quantity, data analysis of sales, etc. thereby to achieve. However, due to the wide range of goods, identification of samples strange and other factors, it is difficult to improve the recognition accuracy of the algorithm, in order to solve this problem, providing a large mass of data for training is a good method, but the shelves of merchandise acquired image data, basically rely on artificial get shot, time-consuming and labor-intensive. This project addresses this shortcoming, we propose a model of commodity-based studio shoot image acquisition and training system to build the library, you can send raspberry programming, selling a variety of merchandise images obtained using the product studio shot mode, to synthesize and simulate shelf commodity image in reality, thereby improving product identification to obtain training data set efficiency and save a lot of unnecessary time and manpower.

Functional Requirements

 

 

Demand priority

functional module	priority
log in	5
User rights management	3
Image upload, search, delete	5
Image tag management	5
Image review	4
graphic gathering	3
Image segmentation	4
Image synthesis	4
System Log		2

Experiences

　　Needs analysis and then draw process requirements document, in fact, a deeper understanding of the project, the process to communicate the depth of the team.

　　Needs analysis is between a bridge between the software systems analysis and design phase. On the one hand, requirements analysis to system specification and project planning as the basic starting point for analysis activities, and the inspection and adjust them from a software point of view; on the other hand, requirements specification and software design, implementation, testing until the main basis for maintenance . Good analytical activities help as soon as possible to avoid or eliminate errors early, thus improving software productivity and reduce development costs, improve software quality.

　　Needs analysis, the entire project is "comminution", "chewed", maneuverability considerations, workload, technical and other factors, put forward detailed requirements, and then extracted to determine the needs of the demand from use cases. The whole process takes everyone into the team, met to discuss several times, many times to communicate with the teacher project, in order to obtain the final result. Project requirements are constantly changing, so every discussion is a result of the remodeling, and is now completed requirements document may not ultimately needs of the project. So to do ongoing discussions, be prepared for continuous change.

　　Example extracted with needs, requirements document is completed, the completion UML use case of FIG. Upon completion of this work, we still encounter many different problems, their relationship to the project in the different roles of authority, functions, relationships between roles. Therefore, throughout the process, the team of all the project depth of analysis, has great significance for the subsequent implementation.