CN-122023141-A - Infrared correction software architecture based on zynq platform and implementation method

CN122023141ACN 122023141 ACN122023141 ACN 122023141ACN-122023141-A

Abstract

The invention discloses an infrared correction software architecture based on zynq platforms and an implementation method thereof, wherein the infrared correction software architecture comprises a PL end and a PS end of the zynq platforms, an infrared detector, an instruction serial port of an upper computer and an image coding chip, wherein the PL end drives the infrared detector to acquire original image data, the PL end receives and buffers the original image data, the instruction serial port transmits serial port instructions to the PL end, the PL end communicates and drives the PS end to carry out correction operation and control data storage/loading in an instruction interrupt mode, the original image data is divided into two paths, one path is transmitted to the PS end to carry out correction operation, a correction operation result is loaded to the other path to obtain an infrared correction image, the correction operation result is configured in the infrared detector, and the infrared correction image data is driven by the PL end and is output to the image coding chip. The invention realizes the infrared correction software architecture based on zynq platform, has smaller power consumption, saves hardware cost, enhances flexibility and expansibility of software, and saves manpower development time cost.

Inventors

ZHANG TIANYI

Assignees

北京遥感设备研究所

Dates

Publication Date: 20260512
Application Date: 20251230

Claims (8)

1. An infrared correction software architecture based on zynq platforms comprises a PL end and a PS end of the zynq platforms and is characterized by further comprising an infrared detector, an instruction serial port of an upper computer and an image coding chip; The PL end is utilized to drive an infrared detector to acquire original image data, and the PL end receives and buffers the original image data; the instruction serial port is used for transmitting serial port instructions to the PL end, and the PL end is used for driving the PS end to carry out correction operation and control data storage/loading in a communication mode of instruction interruption; The original image data obtained through buffering is divided into first path image data and second path image data, the first path image data is transmitted to a PS end to carry out correction operation, a first part of correction operation result is loaded into the second path image data to obtain an infrared correction image, and a second part of correction operation result is configured into an infrared detector; And the infrared correction image data is driven by the PL terminal and is output to the image coding chip.
2. The zynq-platform-based infrared correction software architecture as claimed in claim 1, wherein the PL end comprises a detector driving unit, a detector data buffer unit, a communication driving unit, an image output driving unit and a PL end communication bus, and the PS end comprises a Flash control unit, an OCC coefficient calculation configuration unit, a blind pixel judgment unit and a correction coefficient calculation unit; The detector driving unit is used for driving the infrared detector to acquire original image data; The detector data buffer unit is used for receiving and processing original image data, and the processed original image data is divided into first path image data and second path image data; The communication driving unit is used for receiving the serial port instruction transmitted by the instruction serial port, and communicating and driving the Flash control unit to control data storage/loading in an instruction interrupt mode, and simultaneously communicating and driving the OCC coefficient calculation configuration unit, the blind pixel judgment unit and the correction coefficient calculation unit to operate by utilizing the first path of image data; The PL end communication bus is used for conveying first path image data, correction operation results and infrared correction image data; The blind pixel judging unit is used for judging the position of the blind pixel to form a blind pixel table; the correction coefficient calculation unit is used for calculating a two-point correction coefficient; the blind pixel table and the correction coefficient are used as a first partial correction operation result and are loaded into second path image data to obtain an infrared correction image; The OCC coefficient calculation configuration unit is used for calculating an OCC coefficient as a second partial correction operation result and configuring the OCC coefficient into the infrared detector; The image output driving unit is configured to drive output of the infrared correction image data.
3. The zynq platform-based infrared correction software architecture as claimed in claim 2, wherein the OCC coefficient calculation configuration unit is configured to calculate OCC coefficients, and specifically includes determining OCC coefficients using a binary search method.
4. The zynq platform-based infrared correction software architecture according to claim 2, wherein the correction factor calculation unit is configured to calculate a two-point correction factor, and specifically includes: and calculating a two-point correction coefficient by adopting a non-uniform correction algorithm.
5. The zynq platform-based infrared correction software architecture according to claim 2, wherein the blind pixel determining unit is configured to determine that a blind pixel position forms a blind pixel table, and specifically includes: Judging the blind pixels by a two-point calibration correction method to form a blind pixel table; Or setting reasonable change percentage, and judging the blind pixels to form a blind pixel table by using the normalized response of all array elements to uniform blackbody radiation at high temperature and low temperature.
6. The zynq platform based infrared correction software architecture as set forth in claim 2, wherein the PL-side communication bus is an AXI bus.
7. The zynq platform-based infrared correction software architecture as set forth in claim 1, wherein the instruction serial port is configured to send serial port instructions to the PL, and specifically send serial port instructions using an SPI peripheral bus.
8. The implementation method of the infrared correction software architecture based on zynq platforms is characterized by comprising the following steps: the PL end of zynq platform is used for designing necessary digital input of the infrared detector, comprising an infrared detector working main clock and an integral time signal, driving the infrared detector to acquire image data, decoding and buffering the output data of the infrared detector to obtain original image data, and dividing the buffered original image data into a first path of image data and a second path of image data; the PS end of the zynq platform is informed to respond instruction actions in an instruction interrupt mode, wherein the instruction actions comprise correction operation and control data storage/loading; performing correction operation by using a PS end of the zynq platform, wherein the correction operation comprises correction coefficients, OCC coefficients, judging the positions of blind pixels in an infrared original image to form a blind pixel position vector table, and mobilizing FLASH for data storage; The PL terminal communication bus is utilized to transmit the correction coefficient and the blind pixel position vector table back to the PL terminal, and PL loads the correction coefficient and the blind pixel position vector table on the second path of image data to obtain an infrared correction image; and designing an image output driver by using the PL end of the zynq platform, and outputting the infrared correction image to an image coding chip according to a required image protocol.

Description

Infrared correction software architecture based on zynq platform and implementation method Technical Field The invention belongs to the technical field of infrared thermal imaging, and particularly relates to an infrared correction software architecture based on zynq platforms and an implementation method. Background Zynq is the first expandable processing platform in the industry proposed by the company Xilinx, which is supported by related tools, IP suppliers and other ecosystems, and tightly integrates ARM-Cortex-A9 processor on-chip systems and 28nm low-power programmable logic, thereby helping system architects and embedded software developers expand, customize and optimize systems and providing ideal hardware foundation for constructing efficient heterogeneous computing platforms. In the infrared imaging field, inherent pixel response non-uniformity of an infrared focal plane array can cause fixed pattern noise to be generated in an image, and imaging quality is severely restricted. The existing correction technology is mainly divided into two types, namely a scheme based on special hardware, wherein real-time performance can be guaranteed, algorithm solidification is difficult to adapt to sensor drift and environmental change, and a software scheme based on a general processor is good in flexibility, insufficient in real-time performance and incapable of meeting dynamic scene requirements. Although a hybrid scheme adopting an FPGA or a DSP is adopted to try to improve the performance, the system architecture of the system often cannot realize deep and efficient coordination of software and hardware, and the problems of difficult consideration of instantaneity, flexibility, power consumption and expandability generally exist. Disclosure of Invention Aiming at the problem that the real-time performance, flexibility, power consumption and expandability of the infrared correction software are difficult to consider, the infrared correction software architecture based on the zynq platform is provided by combining the characteristics of the zynq platform. Compared with the traditional single FPGA architecture platform, the hardware resource advantage of the platform is exerted based on the zynq platform, the hardware resource advantage of the platform is fewer, the hardware cost of the platform is lower, the power consumption is lower, the hardware cost is saved, the flexibility and the expansibility of software are enhanced, the compiling speed of an FPGA section is obviously accelerated, and the manpower development time cost is saved. In order to achieve the above purpose, the invention adopts the following technical scheme: in a first aspect, the invention provides an infrared correction software architecture based on zynq platforms, which comprises a PL end and a PS end of the zynq platforms, an infrared detector, an instruction serial port of an upper computer and an image coding chip; The PL end is utilized to drive an infrared detector to acquire original image data, and the PL end receives and buffers the original image data; the instruction serial port is used for transmitting serial port instructions to the PL end, and the PL end is used for driving the PS end to carry out correction operation and control data storage/loading in a communication mode of instruction interruption; The original image data obtained through buffering is divided into first path image data and second path image data, the first path image data is transmitted to a PS end to carry out correction operation, a first part of correction operation result is loaded into the second path image data to obtain an infrared correction image, and a second part of correction operation result is configured into an infrared detector; And the infrared correction image data is driven by the PL terminal and is output to the image coding chip. In some possible implementations, the PL terminal comprises a detector driving unit, a detector data buffer unit, a communication driving unit, an image output driving unit and a PL terminal communication bus, wherein the PS terminal comprises a Flash control unit, an OCC coefficient calculation configuration unit, a blind pixel judgment unit and a correction coefficient calculation unit; The detector driving unit is used for driving the infrared detector to acquire original image data; The detector data buffer unit is used for receiving and processing original image data, and the processed original image data is divided into first path image data and second path image data; The communication driving unit is used for receiving the serial port instruction transmitted by the instruction serial port, and communicating and driving the Flash control unit to control data storage/loading in an instruction interrupt mode, and simultaneously communicating and driving the OCC coefficient calculation configuration unit, the blind pixel judgment unit and the correction coefficient calculation unit to operate by util