A Field-of-View Multi-Fire Source Localization System and Method Based on DSP Image Processing

Abstract

This invention discloses a field-of-view multi-fire source localization system and method based on DSP image processing, relating to the field of fire source image recognition technology. The system includes acquiring multiple consecutive frames of image data, establishing a real-time display loop, storing and directly outputting the first frame image, checking for high grayscale values ​​in the acquisition buffer upon receiving a fire alarm, immediately sending a stop command to the water cannon if high grayscale values ​​are found, solving and storing the corresponding image feature map using the image difference method, using a greedy algorithm to outline the flames, saving all flame outlines greater than a threshold in the field of view when there are multiple flames, and saving the current flame outline for a single flame, calculating the camera coordinates of the largest flame block and the relative coordinates of the other flame blocks to locate the largest flame block, calculating the screen coordinates of the flame block by combining the relative coordinates of the other flame blocks, and selecting the flame block with the closest coordinate distance to begin the next round of scanning for fire extinguishing. This invention helps to achieve the effect of accurately locating and extinguishing fire sources.

Description

Technical Field

[0001] This invention relates to the field of fire source image recognition technology, and in particular to a field-of-view multi-fire source localization system and method based on DSP image processing. Background Technology

[0002] Large-space intelligent fire monitor systems are recognized as highly efficient and rapid fire extinguishing technologies, especially capable of detecting and extinguishing fires in their early stages. They can solve fire protection problems in large-space buildings, achieve fully automated fire protection, and provide strong technical support for security in the construction industry.

[0003] Currently, intelligent fire monitors use methods such as infrared sensors and video image positioning to locate fire sources. Document CN114187364A discloses a rapid fire source location method, which uses infrared thermal imaging technology to record the current position of the image's center of gravity in the field of view, uses a center of gravity algorithm to calculate the coordinates of the fire source's center of gravity on the image, and performs horizontal and vertical calibration on the image to ensure that the fire source's center of gravity is at the center of the field of view, thus completing the location. Document CN105056455B invented a fire monitor location method and device based on image recognition. The processing unit analyzes and processes the image, marks the center of the fire source area on the image, and determines the positioning angle of the fire monitor by adjusting the angle formed by the position of the flame area's center relative to the image center and the monitor tube. Document CN104971458B discloses a multi-fire source identification method based on an automatic tracking and positioning jet extinguishing device. When triggered by a fire warning signal, the video image acquisition module in the extinguishing device is activated to identify and confirm the fire signal. Through target fusion and segmentation, it accurately identifies each flame target and its maximum ignition point.

[0004] Currently, infrared array sensors have a narrow field of view when locating fire sources. In large spaces, it is difficult to effectively solve the coordinates of distant fire sources, and the dynamic processing of fire sources is poor, with significant interference from static flame fluctuations. Existing methods for capturing fire sources using cameras may result in the selection of fire sources without a specified rule in complex situations with multiple flames, leading to low fire extinguishing efficiency.

[0005] Therefore, proposing a field-of-view multi-fire source localization system and method based on DSP image processing to solve the problems existing in the prior art is an urgent problem to be solved by those skilled in the art. Summary of the Invention

[0006] In view of this, the present invention provides a field-of-view multi-fire source localization system and method based on DSP image processing, which can achieve the effect of accurately locating and extinguishing fire sources.

[0007] To achieve the above objectives, the present invention adopts the following technical solution:

[0008] A field-of-view multi-fire source positioning system based on DSP image processing includes an embedded microprocessor DSP, a CCD module, a communication module, a display module, and a water cannon. The CCD module, the embedded microprocessor DSP, and the display module are connected in sequence, and the embedded microprocessor DSP and the water cannon are signal-connected to the communication module.

[0009] The above system, optionally, includes an embedded microprocessor (DSP) comprising an acquisition buffer, a display buffer, a previous frame temporary buffer, an accumulation temporary buffer, a display temporary buffer, and an open frame state buffer.

[0010] A field-of-view multi-fire source localization method based on DSP image processing includes the following steps:

[0011] Data acquisition steps: Acquire image data from multiple consecutive frames acquired by the CCD module to obtain the initial dataset, establish a real-time display loop, store the first frame image in the acquisition buffer and output it directly;

[0012] Flame scanning steps: Upon receiving the fire alarm information from the water cannon, check if there are high grayscale values ​​in the acquisition buffer. If so, immediately send a stop command to the water cannon and store the corresponding image in the previous frame temporary buffer and the accumulated temporary buffer.

[0013] Image processing steps: Use the image difference method to solve the feature map of the previous frame's temporary buffer image and store it in the accumulated temporary buffer. Use a greedy algorithm to outline the position of the flame block in the display temporary buffer.

[0014] Flame frame marking steps: If there are multiple flames, discard the current flame frame if it is too small, calculate and save all flame frames in the field of view that are larger than the threshold; if it is a single flame, save the flame frame.

[0015] Flame frame comparison steps: Compare all flame frames, calculate the camera coordinates of the largest flame block and the relative coordinates of the remaining flame blocks;

[0016] Fire extinguishing procedure: Locate the largest flame block based on camera coordinates. Using the current screen center as a reference point, calculate the screen coordinates of the remaining flame blocks by combining their relative coordinates. Select a flame block with the closest coordinates to begin the next round of scanning and extinguishing.

[0017] Optionally, the image processing steps in the above method may include:

[0018] S1. Obtain multiple frames of data using the initial dataset, perform differential accumulation on the data of consecutive frames, and store them in the accumulation temporary buffer.

[0019] S2: Traverse and display the temporary buffer area, perform a first-line boundary search, find the upper boundary of the flame box of a row of white dots, and initialize the left and right boundaries;

[0020] S3: Use a greedy algorithm to outline the position of the flame block in the temporary display buffer area. Starting from the top boundary, traverse the rows downwards, expand the left and right boundaries, count the number of white dots, determine whether the current row is occupied by white dots, update the bottom boundary, until the number of rows with black dots is greater than the set value.

[0021] Optionally, the above method includes S1: acquiring the second frame image and performing differential processing with the previous frame image, dividing the difference value by a set value and rounding it down, and adding it to the accumulated temporary buffer; repeating the operation until the set frame is read, storing the data in the accumulated temporary buffer into the display temporary buffer, and clearing the accumulated temporary buffer.

[0022] Optionally, in the flame frame marking step, if there are multiple flames, the nearest one should be chosen. When converting from the world coordinate system to the camera coordinate system, the smaller flame frame should be discarded and the larger flame frame should be retained.

[0023] As can be seen from the above technical solutions, compared with the prior art, the present invention provides a field-of-view multi-fire source localization system and method based on DSP image processing, which has the following beneficial effects: 1) The present invention uses the image difference method to process the static flame jitter problem, which can accurately solve the field-of-view fire source coordinates and estimate the target fire source size; 2) The present invention uses a fire source selection strategy suitable for complex fire sources in large spaces, which prioritizes the elimination of fires near the source and fires with large fluctuations in fire source. Attached Figure Description

[0024] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on the provided drawings without creative effort.

[0025] Figure 1 This is a block diagram of a field-of-view multi-fire source localization system based on DSP image processing disclosed in this invention;

[0026] Figure 2 This is a flowchart of a field-of-view multi-fire source localization method based on DSP image processing disclosed in this invention;

[0027] Figure 3 This is a schematic diagram of a field-of-view multi-fire source localization principle based on DSP image processing disclosed in this invention. Detailed Implementation

[0028] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0029] In this application, relational terms such as "first" and "second" are used merely to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. The terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitation, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes said element.

[0030] Reference Figure 1 As shown, the present invention discloses a field-of-view multi-fire source positioning system based on DSP image processing, including an embedded microprocessor DSP, a CCD module, a communication module, a display module and a water cannon. The CCD module, the embedded microprocessor DSP and the display module are connected in sequence, and the embedded microprocessor DSP and the water cannon are signal connected to the communication module.

[0031] Furthermore, the embedded microprocessor DSP includes an acquisition buffer, a display buffer, a previous frame temporary buffer, an accumulation temporary buffer, a display temporary buffer, and an open frame state buffer.

[0032] Furthermore, the CCD module is equipped with a filter, and the water cannon is equipped with an ultraviolet sensor.

[0033] A field-of-view multi-fire source localization method based on DSP image processing, referring to Figure 2 and Figure 3 As shown, it includes the following steps:

[0034] Data acquisition steps: Acquire image data from multiple consecutive frames acquired by the CCD module to obtain the initial dataset, establish a real-time display loop, store the first frame image in the acquisition buffer and output it directly;

[0035] Flame scanning steps: Upon receiving the fire alarm information from the water cannon, check if there are high grayscale values ​​in the acquisition buffer. If so, immediately send a stop command to the water cannon and store the corresponding image in the previous frame temporary buffer and the accumulated temporary buffer.

[0036] Image processing steps: Use the image difference method to solve the feature map of the previous frame's temporary buffer image and store it in the accumulated temporary buffer. Use a greedy algorithm to outline the position of the flame block in the display temporary buffer.

[0037] Flame frame marking steps: If there are multiple flames, discard the current flame frame if it is too small, calculate and save all flame frames in the field of view that are larger than the threshold; if it is a single flame, save the flame frame.

[0038] Flame frame comparison steps: Compare all flame frames, calculate the camera coordinates of the largest flame block and the relative coordinates of the remaining flame blocks;

[0039] Fire extinguishing procedure: Locate the largest flame block based on camera coordinates. Using the current screen center as a reference point, calculate the screen coordinates of the remaining flame blocks by combining their relative coordinates. Select a flame block with the closest coordinates to begin the next round of scanning and extinguishing.

[0040] Furthermore, in the image data acquisition step, the image data acquired by the CCD module includes: each pixel of the CCD image with a filter is an 8-bit grayscale value; in order to effectively reduce the influence of radial distortion and improve speed, the detection field of view can be reduced inward when matching different application environments.

[0041] Furthermore, in the flame scanning step, obtaining the fire alarm information of the water cannon includes: the water cannon judging whether a flame is detected in the large space; if a fire occurs, it immediately communicates with the DSP via serial port, the water cannon performs gradient search, and the DSP performs image processing and recognition; when a high grayscale value is detected in the image, the water cannon immediately stops searching, stores this frame image in the previous frame temporary buffer area and stores it in the accumulated temporary buffer area as the initial value.

[0042] Further image processing steps include:

[0043] S1. Obtain multiple frames of data using the initial dataset, perform differential accumulation on the data of consecutive frames, and store them in the accumulation temporary buffer.

[0044] S2: Traverse and display the temporary buffer area, perform a first-line boundary search, find the upper boundary of the flame box of a row of white dots, and initialize the left and right boundaries;

[0045] S3: Use a greedy algorithm to outline the position of the flame block in the temporary display buffer area. Starting from the top boundary, traverse the rows downwards, expand the left and right boundaries, count the number of white dots, determine whether the current row is occupied by white dots, update the bottom boundary, until the number of rows with black dots is greater than the set value.

[0046] Furthermore, S1 includes: acquiring the second frame image and performing differential processing with the previous frame image, dividing the difference value by a set value and rounding it down, and adding it to the accumulated temporary buffer; repeating the operation until the set frame is read, storing the data in the accumulated temporary buffer into the display temporary buffer, and clearing the accumulated temporary buffer.

[0047] Furthermore, in S2, white dots are pixels with gray values ​​not less than a set threshold, and their search state has not been transferred; they remain in the search state. A row of white dots is defined as a row with a continuous number of white dots that is greater than a set value.

[0048] Specifically, the definition of continuous white dot count is that the number of black dots between two white dots is no greater than a set number. Once the number of black dots exceeds the set number, the flame frame ends at the right boundary of the current row. Black dots are pixels with gray values ​​less than a set threshold.

[0049] Furthermore, the greedy algorithm in S3 for outlining the flame block includes:

[0050] S3.1. Starting from the left boundary of the flame frame, traverse to the left; if a white dot is found, update the left boundary; otherwise, update the width of the black dot until the number of black dots is greater than the set value, then exit the traversal to the left.

[0051] S3.2: Starting from the right boundary of the flame frame, traverse to the right; if a white dot is found, update the right boundary; otherwise, update the width of the black dot, until the number of black dots is greater than the set value, then exit the traversal to the right;

[0052] S3.3: Determine if the current row is a white dot row. If it is, update the lower boundary. Otherwise, increment the height of the black dot row by 1. When the number of black dot rows exceeds the set value, exit the traversal and find the flame box to be evaluated.

[0053] Furthermore, in the flame frame marking step, when there are multiple flames, the nearest one is adopted. When converting from the world coordinate system to the camera coordinate system, the smaller flame frame is discarded and the larger flame frame is retained.

[0054] Specifically, the threshold values ​​in the selection criteria for the flame frame can be flexibly switched according to the user or application scenario.

[0055] Furthermore, in the flame frame marking step, the flame block coordinates are the coordinates of the center point of the flame frame.

[0056] The various embodiments in this specification are described in a progressive manner. Similar or identical parts between embodiments can be referred to mutually. Each embodiment focuses on describing the differences from other embodiments. In particular, for system or system embodiments, since they are basically similar to method embodiments, the description is relatively simple, and relevant parts can be referred to the descriptions in the method embodiments. The systems and system embodiments described above are merely illustrative. The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs. Those skilled in the art can understand and implement this without creative effort.

[0057] The above description of the disclosed embodiments enables those skilled in the art to make or use the invention. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the invention. Therefore, the invention is not to be limited to the embodiments shown herein, but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A field of view multi-source positioning method based on DSP image processing, characterized in that, It comprises the following steps: Data acquisition step: collect the image data obtained by the continuous multiple frames of CCD module to obtain the initial data set, establish the real-time cycle of display, store the first frame image into the acquisition buffer and directly output; Flame scanning step: receive the fire alarm information of the water cannon, check whether there is high gray value in the acquisition buffer, if yes, immediately send the stop command to the water cannon, store the corresponding image into the last frame temporary buffer and the accumulated temporary buffer; Image processing step: solve the feature image in the last frame temporary buffer image using image difference method and store it into the accumulated temporary buffer, use the greedy algorithm to frame the position of the flame block in the display temporary buffer; Flame frame marking step: when there are multiple flames, if the current flame frame is small, discard it, calculate and save all flame frames in the field of view which are greater than the threshold value; If it is a single flame, save the current flame frame; Flame frame comparison step: compare all flame frames, calculate the camera coordinates of the largest flame block and the relative coordinates of the remaining flame blocks; Positioning and extinguishing step: according to the camera coordinates, position the largest flame block, take the current screen center as the reference point, combine the relative coordinates of the remaining flame blocks to calculate the screen coordinates of the remaining flame blocks, select a flame block with close distance to start the next round of scanning and extinguishing; The image processing step comprises: S1, using the initial data set to obtain multiple frames of data, performing difference accumulation processing on the front and back frames of data, and storing into the accumulated temporary buffer; S2: traverse the display temporary buffer, perform first row boundary search, find the upper boundary of a row of white point flame frame, and initialize the left and right boundaries; S3: using the greedy algorithm to frame the position of the flame block in the display temporary buffer, starting from the upper boundary, traversing the row downward, expanding the left and right boundaries, counting the number of white points, judging whether the current row is occupied by white points, updating the lower boundary, and stopping until the number of black points is greater than the set value.

2. The field of view multi-fire source positioning method based on DSP image processing according to claim 1, wherein S1 comprises: collecting the second frame image, and performing difference processing on the second frame image and the previous frame image, dividing the difference value by a set value, taking the integer, and adding the integer to the accumulated temporary buffer; repeating the operation until a set number of frames are read, storing the data in the accumulated temporary buffer into the display temporary buffer, and emptying the accumulated temporary buffer.

3. The field of view multi-fire source positioning method based on DSP image processing according to claim 1, wherein in the flame frame marking step, when there are multiple flames, the nearest principle is adopted, when the nearest in the world coordinate system is converted into the camera coordinate system, the small flame frame is discarded and the large flame frame is retained. It comprises an embedded microprocessor DSP, a CCD module, a communication module, a display module and a water cannon, the CCD module, the embedded microprocessor DSP and the display module are connected in sequence, and the embedded microprocessor DSP and the water cannon are signal connected with the communication module.

5. The field of view multi-fire source positioning system based on DSP image processing according to claim 4, wherein the embedded microprocessor DSP comprises an acquisition buffer, a display buffer, a last frame temporary buffer, an accumulated temporary buffer, a display temporary buffer and an open frame state buffer.

4. A DSP image processing based field of view multi-fire source location system for implementing the DSP image processing based field of view multi-fire source location method of any one of claims 1-3, characterized in that, ​ ​ ​

Images