System solution for fingerprint recognition using DSP high-speed processing capability

1 Introduction

Fingerprint recognition technology analyzes the local features of fingerprints and extracts detailed feature points from them, thereby reliably confirming personal identity. Fingerprint recognition not only has many unique information security advantages, but more importantly, it is highly practical and feasible.

At present, most fingerprint recognition systems collect fingerprint images into computers and use computers for recognition. Independent fingerprint recognition systems produced by some foreign companies are relatively expensive. All these have limited the popularization of fingerprint recognition technology. Therefore, researching and developing fast, high-recognition, and low-cost independent fingerprint recognition systems has great market prospects and important scientific research value.

This paper proposes a new fingerprint recognition system based on DSP. In terms of hardware, it uses the high-speed processing capability of DSP to build a high-speed data processing platform. In terms of software, it refers to the processing characteristics of DSP and hardware logic to improve the traditional fingerprint algorithm to meet the real-time and reliability requirements.

2 Hardware System Structure

The principle block diagram of the system is shown in Figure (1):

Figure (1) System structure diagram

2.1 Image acquisition module

In the image acquisition module, since the fingerprint recognition system does not require real-time image observation, the requirements for the sensor are not very high, and general black and white digital CMOS sensors can meet the requirements. This system uses a 3-megapixel high-definition black and white sensor as an image acquisition device, which is very suitable for use as a fingerprint image sensor. The main considerations are the advantages of CMOS devices, such as low cost, high resolution, and good reliability. The disadvantage is that the image quality may deteriorate when the fingers are sweaty or cracked. In the image recognition process, an enhancement algorithm based on GABOR is used, which can basically overcome the resulting impact. 2.2 Image processing and recognition module

The structure of the image processing and recognition module is related to the overall performance of the system. The FPGA+ DSP architecture is conducive to building an efficient data processing process and facilitating the allocation of processing tasks, thereby improving the system's parallelism and resource utilization. The SRAM, SDRAM, and FLASH in the system are directly connected to the DSP for use: FLASH is used to store programs and some fixed table data; SDRAM is the system memory of the DSP and is used for the operation of system programs; SRAM is a high-speed data storage area used to store temporary variables generated when the program is running. The DDR SDRAM is specifically used to store some large-capacity data blocks such as the collected fingerprint data and the pixel gradient data calculated during the preprocessing process. It is directly connected to the FPGA and is the highest-speed memory area in the system. In addition to being the expansion bus interface of the DSP processor, the FPGA also shares some data processing tasks, because a single DSP cannot handle all calculation and control tasks. When processing fingerprint data, some cumbersome addition and subtraction operations and logic operations are often encountered. Usually, this part is handled by the FPGA. Considering the particularity of the fingerprint processing algorithm, the DDR control function must also be realized.

Due to the large amount of mathematical calculations in the fingerprint recognition process, the program design inevitably requires a large storage space. In order to improve the overall performance, the heavy calculation tasks need to be handed over to the DSP for processing, while the image acquisition part should occupy as little DSP time as possible. In addition, using the gaps in image acquisition, or while the image is being acquired, the hardware can complete some simple and tedious calculations to share the DSP's processing tasks, improve the parallelism of processing, and meet the requirements of real-time performance. This system uses TMS320VC5402, which has a fast computing speed and a high cost-effectiveness. The 8-bit grayscale fingerprint image collected in the system occupies one byte per pixel, and the image size is 512×512 pixels. It takes 256k bytes of storage space to store one frame of the image. The DSP unit is the core of the entire fingerprint processing system and is responsible for real-time processing of fingerprints.

2.3 Output Module

As an independent fingerprint recognition system, the data recognized by the system can be directly displayed on the LCD. When designing the system, the system can also be used as a terminal, that is, the Ethernet interface is expanded through FPGA, as a large fingerprint recognition system terminal that needs to transmit fingerprint library data through the network.

3 Fingerprint Recognition Algorithm

The fingerprint recognition algorithm is the core of fingerprint recognition. The fingerprint recognition algorithm process used in this system is shown in Figure (2).

Figure (2) Fingerprint recognition algorithm flow

Figure (2) Fingerprint recognition algorithm flow

Except for the singular area, the texture of the fingerprint image in a sufficiently small area is similar to parallel straight lines, which is the directional feature of the fingerprint image. The directional feature is one of the most obvious features of the fingerprint image. It intuitively reflects the basic morphological features of the fingerprint image in a simplified form, and is therefore widely used in fingerprint image classification, enhancement, feature extraction, and other aspects.

The method for extracting the ridge direction is:

⑴ Divide the fingerprint image into sub-blocks that are small enough to satisfy the condition that the textures in the blocks are approximately parallel.