README.md 11.3 KB
Newer Older
Sabyasachi Mondal's avatar
Sabyasachi Mondal committed
1
# FPGA in Image Feature detection and comparison with CPU
Sabyasachi Mondal's avatar
Sabyasachi Mondal committed
2
#### Group4: Sabyasachi Mondal , Ravi Yadav
Sabyasachi Mondal's avatar
Sabyasachi Mondal committed
3
fpga for streamlining of computation intensive tasks. In this case we take an hyperspectral image which is generally analysed by satellites or drones mostly consisting of single band image data. This can be used for both maritime and vehicular navigation.
Sabyasachi Mondal's avatar
Sabyasachi Mondal committed
4

Sabyasachi Mondal's avatar
Sabyasachi Mondal committed
5
6
7
8
9
10
11
12
13
14
15
#### I Have tried to build a implementation from scratch where all code structure were created so we can have much higher control over design
| Title | Content |
| ------ | ------ |
| Overview | Why do we use FPGA |
| Problem background | Where can we apply our methods and get results |
| Objective | Our vision of the solution |
| Implementation | The Hardware architechture and HLS code constructs |
| Robert's cross | The algorithm we implemented |
| What we achieved and Caveats | Objectives sucessful and limitations |
| Future Scope | Possible use in multiagent robot control using continous streaming |

Sabyasachi Mondal's avatar
Sabyasachi Mondal committed
16
## Overview
Sabyasachi Mondal's avatar
Sabyasachi Mondal committed
17
18
19
We want to use FPGA for implementing an algorithm in hardware to perform computation more effeciently. CPU hardware is non-flexible so the code runs using the same set of registers and ALU , we cant optimize the harware as per our code. Our objective here is to harware a processing unit (something smilar to a flexible ALU using the CLBs) in the FPGA using High level code.


Sabyasachi Mondal's avatar
Sabyasachi Mondal committed
20
21
## Problem background
<b>*For applications like real time image processing, using CPU resource can be expensive, and the reaction time may increase in applications where decisions are based on calculations. We need a dedicated hardware that can process continous process of data coming in from sensors or camera endlessly.*</b>
Sabyasachi Mondal's avatar
Sabyasachi Mondal committed
22

Sabyasachi Mondal's avatar
Sabyasachi Mondal committed
23
<b>*FPGA should be able to process multiple streams in synchronized manner. We want to process the streams coming from an image and process them through a convolution algorithm (Robert's matrix) and then use another function to filter out relevant parts*</b>
Sabyasachi Mondal's avatar
Sabyasachi Mondal committed
24

Sabyasachi Mondal's avatar
Sabyasachi Mondal committed
25
## Objective
Sabyasachi Mondal's avatar
Sabyasachi Mondal committed
26

Sabyasachi Mondal's avatar
Sabyasachi Mondal committed
27
Our objective is to use enable continous data stream processing in a pipeline that runs faster using FPGA in comparison to CPU.
Sabyasachi Mondal's avatar
Sabyasachi Mondal committed
28
We try to implement a image-filter which works by taking data streams and processing them on real time, and the FPGA should work faster than CPU. Our objective is not to make the image-processing-algorithm fast.
Sabyasachi Mondal's avatar
Sabyasachi Mondal committed
29

Sabyasachi Mondal's avatar
Sabyasachi Mondal committed
30
We should be able to:
Sabyasachi Mondal's avatar
Sabyasachi Mondal committed
31

Sabyasachi Mondal's avatar
Sabyasachi Mondal committed
32
1. <b>*Remove limitations on length and size of data so the design can be adapted for real-time continous use with streaming data*</b>
33
    
Sabyasachi Mondal's avatar
Sabyasachi Mondal committed
34
2. <b>*Enable multiple data stream processing is parallel using the strategies used in HLS for faster processing*</b>
Sabyasachi Mondal's avatar
Sabyasachi Mondal committed
35
36

3. <b>*FPGA should be reasonably faster than our CPU for processing streams*</b>
Sabyasachi Mondal's avatar
Sabyasachi Mondal committed
37

Sabyasachi Mondal's avatar
Sabyasachi Mondal committed
38
## Implementation Strategy
Sabyasachi Mondal's avatar
Sabyasachi Mondal committed
39
Previously we have seen the image resizer takes in the whole data DMA makes the data transfer rate much faster, but we cant pr``ocess an image or stream of data that is infinitely received and require processing.
Sabyasachi Mondal's avatar
Sabyasachi Mondal committed
40
41

We intend to implement the following:
Sabyasachi Mondal's avatar
Sabyasachi Mondal committed
42
1. <b>*fast multichannel stream operations at a hardware level integrated with similar high level software constructs in python*</b>
43
    
Sabyasachi Mondal's avatar
Sabyasachi Mondal committed
44
    *1.a High Level Code structure in pythonto enable parallel operation and feed data continously using DMA*
45
    
Sabyasachi Mondal's avatar
Sabyasachi Mondal committed
46
47
    *1.b Maintain same level of parallelism (multiple processing streams) in unrolled loops*

Sabyasachi Mondal's avatar
Sabyasachi Mondal committed
48
2. <b>*make the FPGA capable to process continous stream of data which is infeasible to be stored*</b>
Sabyasachi Mondal's avatar
Sabyasachi Mondal committed
49
50

    *2.a CPU packs data and feed them to FPGA till the image is processed (but we can simply loop it forever for continous data)*
51

Sabyasachi Mondal's avatar
Sabyasachi Mondal committed
52
    *2.b Synchronized operation between packets of each stream which is essential for processing multiple togather.*
53

Sabyasachi Mondal's avatar
Sabyasachi Mondal committed
54
We try to read each row in the image as a pack of 3 streams process it in 2 seperate block and return the output as an array. *Since Robert's cross works on a specific 2x2 set of pixels of the image we must synchronize the stream of data coming in from our DMA such that if we read Nth packet in DMA A we must read Nth packet in DMA B and Nth packet in DMA C.*
Sabyasachi Mondal's avatar
Sabyasachi Mondal committed
55

Sabyasachi Mondal's avatar
Sabyasachi Mondal committed
56
![Schematic streaming rows and output](https://mygit.th-deg.de/sm11312/fpga_final_project/-/raw/main/HLSolution.JPG "Schematic streaming rows and output")
57

Sabyasachi Mondal's avatar
Sabyasachi Mondal committed
58
This would mean we can store real-time data in frames and feed them continously from our python code. The processing blocks consists of a 2x2 array each and they are the convolution weights added to our stream of data and we return the output with a pre-processing in another function.
Sabyasachi Mondal's avatar
Sabyasachi Mondal committed
59

Sabyasachi Mondal's avatar
Sabyasachi Mondal committed
60
![Convolution on streaming row](https://mygit.th-deg.de/sm11312/fpga_final_project/-/raw/main/RobertCross.JPG "Convolution on streaming row")
Sabyasachi Mondal's avatar
Sabyasachi Mondal committed
61

Sabyasachi Mondal's avatar
Sabyasachi Mondal committed
62
(DMA1 + DMA2) streams are processed in PU1 and (DMA2 + DMA3) streams in PU2. However becuase Robert's convolution algorithm needs data to be processed in a 2x2 array they must enter and get processed in Synchronized manner.
Sabyasachi Mondal's avatar
Sabyasachi Mondal committed
63
64

![CPU FPGA interconnection and data transfer](https://mygit.th-deg.de/sm11312/fpga_final_project/-/raw/main/CPU_FPGA.JPG "CPU FPGA interconnection and data transfer")
Sabyasachi Mondal's avatar
Sabyasachi Mondal committed
65

Sabyasachi Mondal's avatar
Sabyasachi Mondal committed
66
On the higher level the interaction between CPU and FPGA looks like the schematic shown above.
Sabyasachi Mondal's avatar
Sabyasachi Mondal committed
67

Sabyasachi Mondal's avatar
Sabyasachi Mondal committed
68
We use two blocks to process the streams but that doesnot mean we use one thread we basically dont wait for Nth set of data to be processed before we can start processing N+1 data. Since the convolution algorithm does not wait for processing it can start to read and process the next N+1 set of data from the stream as soon as Nth set has been read. It looks something like this due to loop unrolling and leads to parallel processing.
Sabyasachi Mondal's avatar
Sabyasachi Mondal committed
69
70

![Unravelling of streams in loop and parallel processing](https://mygit.th-deg.de/sm11312/fpga_final_project/-/raw/main/Parallel_process.JPG "Unravelling of streams in loop and parallel processing")
Sabyasachi Mondal's avatar
Sabyasachi Mondal committed
71

Sabyasachi Mondal's avatar
Sabyasachi Mondal committed
72
73
## Instroduction to the algorithm
<br>*The RObert's operator is very effective in detecting features in an image, specifically for images with more precise features*</br>
Sabyasachi Mondal's avatar
Sabyasachi Mondal committed
74

Sabyasachi Mondal's avatar
Sabyasachi Mondal committed
75
<br>*The Robert's operator is a 2x2 Matrix that can be used to find out differences in image bouundaries becuase of it's weights*<br>
Sabyasachi Mondal's avatar
Sabyasachi Mondal committed
76

Sabyasachi Mondal's avatar
Sabyasachi Mondal committed
77
![Introduction to Robert's cross operation](https://homepages.inf.ed.ac.uk/rbf/HIPR2/roberts.htm "Brief introduction to Robert's cross operation")
Sabyasachi Mondal's avatar
Sabyasachi Mondal committed
78

Sabyasachi Mondal's avatar
Sabyasachi Mondal committed
79
## What we achieved and the caveat :
Sabyasachi Mondal's avatar
Sabyasachi Mondal committed
80
<b>*We intended to build a architechture that can process multiple streams and process them in same parallel level and we were sucessful.*</b>
Sabyasachi Mondal's avatar
Sabyasachi Mondal committed
81

Sabyasachi Mondal's avatar
Sabyasachi Mondal committed
82
<b>*Our main goal is to ensure such a architechture runs faster in FPGA and it was reasonably fast; most importantly it can be scaled up to handle multiple streams.*</b>
Sabyasachi Mondal's avatar
Sabyasachi Mondal committed
83
84
85

It is not very suitable for image processing tasks as arrays stored in memory does a better work in that, so a Robert's convolution algorithm is faster in an OpenCV library.

Sabyasachi Mondal's avatar
Sabyasachi Mondal committed
86
<b>*CPU Average for images was at 10s and FPGA at about 6s*</b>
Sabyasachi Mondal's avatar
Sabyasachi Mondal committed
87

Sabyasachi Mondal's avatar
Sabyasachi Mondal committed
88
89
![Speed comparison in single images](https://mygit.th-deg.de/sm11312/fpga_final_project/-/raw/main/Speed_table.JPG "Speed comparison in single images")

Sabyasachi Mondal's avatar
Sabyasachi Mondal committed
90
## Future scope
Sabyasachi Mondal's avatar
Sabyasachi Mondal committed
91
92
*This is a new idea and has no previous references except implementaton guides.*

Sabyasachi Mondal's avatar
Sabyasachi Mondal committed
93
<b>*The image processing can serve a stepping stone for controlling multi-agent systems. Where each streaming interface can be used for instruction input and output for each agent/bots.*</b>
Sabyasachi Mondal's avatar
Sabyasachi Mondal committed
94

Sabyasachi Mondal's avatar
Sabyasachi Mondal committed
95
*We achieved good synchronization betwenn the input streams in terms of pixel processing. We can consider extending the filter to video with streaming, it might be possible with similar kind of streaming interface.*
Sabyasachi Mondal's avatar
Sabyasachi Mondal committed
96
97


Sabyasachi Mondal's avatar
Sabyasachi Mondal committed
98
#### Tasks
Sabyasachi Mondal's avatar
Sabyasachi Mondal committed
99
The Tasks and maximum actual time:
Sabyasachi Mondal's avatar
Sabyasachi Mondal committed
100

Sabyasachi Mondal's avatar
Sabyasachi Mondal committed
101
1. Problem statement and brainstorming for project selection : *24 hrs*
Sabyasachi Mondal's avatar
Sabyasachi Mondal committed
102
103
104
105
106
107
108
2. Design a basic model and build overlay : *6 hrs*
3. Python code adjustment and integration : *4 hrs*
4. Plan next stage of overlay design : *4 hrs*
5. Develop algorithm for FPGA using C++ : *4 hrs*
6. Optimize code and add synchronization of multiple channels : *24 hrs*
7. Implement block diagram : *4 hrs*
8. Upload code and test in IPy notebook : *3 hrs* 
Sabyasachi Mondal's avatar
Sabyasachi Mondal committed
109
110


Sabyasachi Mondal's avatar
Sabyasachi Mondal committed
111
#### Resources used and Future project topics
Sabyasachi Mondal's avatar
Sabyasachi Mondal committed
112

Sabyasachi Mondal's avatar
Sabyasachi Mondal committed
113
##### Resources used
Sabyasachi Mondal's avatar
Sabyasachi Mondal committed
114
0. Images: https://serc.carleton.edu/earth_analysis/image_analysis/introduction/day_4_part_2.html
Sabyasachi Mondal's avatar
Sabyasachi Mondal committed
115
116
117
118
119
120
121
122
123
124
1. Image segmentation : https://theailearner.com/2020/11/29/image-segmentation-with-watershed-algorithm/
2. Operation with stream: https://www.xilinx.com/html_docs/xilinx2020_2/vitis_doc/hls_stream_library.html#ivv1539734234667__ad398476
3. Stream Interface : https://www.xilinx.com/html_docs/xilinx2020_2/vitis_doc/managing_interface_synthesis.html#ariaid-title32
3. Specialized Constructs : https://www.xilinx.com/html_docs/xilinx2020_2/vitis_doc/special_graph_constructs.html?hl=template
4. Vitis Examples : https://github.com/Xilinx/Vitis_Accel_Examples/blob/master/cpp_kernels/README.md
5. Running Accelerator : https://pynq.readthedocs.io/en/v2.6.1/pynq_alveo.html#running-accelerators
6. Pragma Interfaces : https://www.xilinx.com/html_docs/xilinx2017_4/sdaccel_doc/jit1504034365862.html
7. AXI4 : https://ch.mathworks.com/help/hdlcoder/ug/getting-started-with-axi4-stream-interface-in-zynq-workflow.html
8. Interface of Streaming : https://www.xilinx.com/html_docs/xilinx2020_2/vitis_doc/managing_interface_synthesis.html#ariaid-title34
9. Database in FPGA : https://dspace.mit.edu/bitstream/handle/1721.1/91829/894228451-MIT.pdf, https://www.xilinx.com/publications/events/developer-forum/2018-frankfurt/accelerating-databases-with-fpgas.pdf, https://www.xilinx.com/html_docs/xilinx2020_2/vitis_doc/vitis_hls_process.html#djn1584047476918
Sabyasachi Mondal's avatar
Sabyasachi Mondal committed
125
10. Muxed Stream : https://liu.diva-portal.org/smash/get/diva2:1057270/FULLTEXT01.pdf
Sabyasachi Mondal's avatar
Sabyasachi Mondal committed
126
127
11. RAW,WAR,WAW.. : https://www.xilinx.com/html_docs/xilinx2020_2/vitis_doc/vitis_hls_optimization_techniques.html#wen1539734225565__aa1299615
12: Loop Pipelining Roll Unroll : https://www.xilinx.com/html_docs/xilinx2020_2/vitis_doc/vitis_hls_optimization_techniques.html#kcq1539734224846
Sabyasachi Mondal's avatar
Sabyasachi Mondal committed
128

Sabyasachi Mondal's avatar
Sabyasachi Mondal committed
129

Sabyasachi Mondal's avatar
Sabyasachi Mondal committed
130
### Errors Logs and Issues encountered
Sabyasachi Mondal's avatar
Sabyasachi Mondal committed
131
The input pins (listed below) are either not connected or do not have a source port, and they don't have a tie-off specified. These pins are tied-off to all 0's to avoid error in Implementation flow.
Sabyasachi Mondal's avatar
Sabyasachi Mondal committed
132
133
134
Please check your design and connect them as needed: 
/color_filter/ap_start
When ap_Ctrl = None not specified in design
Sabyasachi Mondal's avatar
Sabyasachi Mondal committed
135
136

Cant find custom IP in Vivado : add IP zip path, open IP Integrator view, from IP configure window manually add the IP
Sabyasachi Mondal's avatar
Sabyasachi Mondal committed
137
138
139

Cant connect hls::stream<> type object in IP : Note: The hls::stream class should always be passed between functions as a C++ reference argument. For example, &my_stream.
IMPORTANT: The hls::stream class is only used in C++ designs. Array of streams is not supported.
Sabyasachi Mondal's avatar
Sabyasachi Mondal committed
140
141

Non-Blocking write not-allowed in Non-FIFO Interfaces like axis instead try using FIFO m_axi
Sabyasachi Mondal's avatar
Sabyasachi Mondal committed
142
143

DMA size must be lesser than 16383 so we cant feed very large datasets directly to a single DMA.
Sabyasachi Mondal's avatar
Sabyasachi Mondal committed
144
145
146

WARNING: [HLS 200-786] Detected dataflow-on-top in function  'color_filter' (../project_3/color_filter.cpp:45)  with default interface mode 'ap_ctrl_hs'. Overlapped execution of successive kernel calls will not happen unless interface mode 'ap_ctrl_chain' is used (or 'ap_ctrl_none' for a purely data-driven design).
Resolution: For help on HLS 200-786 see www.xilinx.com/cgi-bin/docs/rdoc?v=2020.2;t=hls+guidance;d=200-786.html
Sabyasachi Mondal's avatar
Sabyasachi Mondal committed
147

Sabyasachi Mondal's avatar
Sabyasachi Mondal committed
148
DMA Stuck and not reponding, [fixed thanks to Lauri's Blog](https://lauri.võsandi.com/hdl/zynq/xilinx-dma.html) and [problems other's faced](https://forums.xilinx.com/t5/Processor-System-Design-and-AXI/Why-AXI-DMA-starts-acquiring-data-during-configuration/td-p/766605) and more [problems](https://forums.xilinx.com/t5/AXI-Infrastructure-Archive/tkeep-signal-in-AXI-DMA-and-tstrb-3-0-in-Custom-AXI-Stream-IP/td-p/921850)