Biometric identification technology refers to the use of the inherent physiological or behavioral characteristics of the human body for personal identification. Some representative physiological features include fingerprint, palm shape, face shape, iris, retina, sound, etc., but iris recognition, face recognition, sound recognition, fingerprint recognition, etc. there are some problems, such as expensive equipment, greatly affected by the environment, easy to forge, special people and so on. Palm shape recognition technology is a new biometric technology, which can well solve the bottleneck of the existing biometric technology. The palm data acquisition circuit discussed in this design provides accurate and reliable original data for this new recognition technology.

  Working principle of data acquisition system

The photosensitive unit is excited by the incandescent light adjusted by brightness braking, converts the optical image of the human palm into an electrical signal output serially according to the sequence, and outputs the analog image signal serially to the front end of the preamplifier circuit under the action of the driving pulse sent by the pulse driving circuit. The analog image signal directly amplified by the preamplifier circuit is output to the A / D converter inside the tms320c5509 DSP processor for a / D conversion after further noise reduction by the filter circuit. The DSP processor stores the converted digital image signal into the external expansion SDRAM, and in the acquisition process, the DSP processor starts to preprocess it to obtain the palm contour, then obtains the effective feature parameters from the processed image according to the structural algorithm of the feature extraction classifier, and finally compares it with the previously collected sample information to complete the acquisition and recognition process.

  Extended and sequential memory circuit design

The size of tcd1206 linear CCD photosensitive unit is: 14mm & tips; 14mm &TImes; 14mm; If 1000 rows of data are collected (the corresponding palm size is about 30.24cm & tips; 14cm, larger than the longest hand in the world 29.6 CM & tips; 12cm), then the corresponding data amount = 2106 × 1000 bytes, that is, the maximum amount of data to be processed by the system for each identification is about 2m bytes. In order to temporarily store such a large amount of data in time and quickly. Hy57v641620hg SDRAM with data width of 16 bits, internal divided into 4 banks and total storage capacity of 4m bytes is used as palm data memory. The extended SDRAM is shown in the figure.

  Circuit design of fingerprint identification data acquisition system

Figure 3 SDRAM memory expansion block diagram

  Design of CCD pre-processing circuit

The DC output signal of the selected CCD is about 4.1V and the maximum saturation output is 450mv. Therefore, in order to facilitate the processing of correlation double sampling and digital to analog conversion in the later stage, the signal must be subjected to pre-processing such as blocking and amplification. The circuit of the pre-processing is shown in the figure. Q1, Q2 and Q3 are emitter followers, and U2, R3 and R4 constitute a preamplifier circuit with a magnification of 21 times; C1 is the DC isolation capacitance; C2 limits the bandwidth of the preamplifier circuit to filter out high-frequency noise and spikes in the signal; The output end is connected in series with a few ohm resistance to match the impedance of the later stage circuit.

The output impedance of the follower is obtained from the formula, where RS is the CCD output resistance 250W, and Rb resistance, that is, R2, B and RBE in Fig. 4 are the circuit magnification and volume resistance of the three-stage transistor respectively.

  Circuit design of fingerprint identification data acquisition system

Fig. 4 front stage processing circuit diagram

RO≈(RS//RB)+rbe/b (1)

Q1 selects the triode mat-02 of ADI company with low noise, low offset and high current magnification. When IC is as low as 1mA, its current magnification is B “500 and the body resistance is 0.3w. According to equation (1), the output impedance of the emitter follower circuit is less than 1W. The preamplifier in the circuit selects the low noise current feedback operational amplifier ad844, the voltage noise density en is 2nv /, and the current noise density in at the in-phase terminal is 12Pa /.

Leave a Reply

Your email address will not be published. Required fields are marked *