一般的读取场的每个时刻的状态，然后输出成视频的 matlab 代码类似这样

video_name = 'zeta_video';
obj = VideoWriter(video_name);
obj.FrameRate = 10;
open(obj);T_STEP = 1.8e-3;
MAX_STEP = 50000;
OUTPUT_GAP = 500;x = linspace(0, 8, 513);
y = linspace(0, 8, 513);
[X,Y] = meshgrid(x,y);levels = linspace(-15, 15, 100);t_len = MAX_STEP/OUTPUT_GAP;n = 0:OUTPUT_GAP:MAX_STEP-OUTPUT_GAP;nImages = length(n);
fig = figure;
%Draw multiple images to display and store each image into a cell array%
for idx = 1:nImageszeta = csvread(sprintf('./zeta/zeta_%d.csv', n(idx)));[C,h] = contourf(X, Y, zeta', levels);clim([-15 15]);xlim([1 3]);ylim([3 5]);colorbar;set(h,'LineColor','none');title(sprintf('zeta t = %.2fs', n(idx)*T_STEP));drawnowframe = getframe(fig); writeVideo(obj,frame);
endclose(obj);

后来我想用 scatter 画的时候，发现帧与帧之间的图像重合了

video_name = 'p_video';
obj = VideoWriter(video_name);
obj.FrameRate = 20;
open(obj);T_STEP = 1.8e-3;
MAX_STEP = 50000;
OUTPUT_GAP = 200;x = linspace(0, 8, 512);
y = linspace(0, 8, 512);
[X,Y] = meshgrid(x,y);levels = linspace(-0.001, 0.001, 100);t_len = MAX_STEP/OUTPUT_GAP;n = OUTPUT_GAP:OUTPUT_GAP:MAX_STEP-OUTPUT_GAP;nImages = length(n);
fig = figure;
%Draw multiple images to display and store each image into a cell array%
for idx = 1:nImagesp = csvread(sprintf('./p/p_%d.csv', n(idx)));p = p';ax = axes();scatter(X(:), Y(:), [], p(:), 'filled');colormap(ax, jet(numel(levels)-1)); % set colormapcaxis(levels([1,end]));             % set color rangecb = colorbar();xlim([1 3]);ylim([3 5]);title(sprintf('zeta t = %.2fs', n(idx)*T_STEP));drawnowframe = getframe(fig); writeVideo(obj,frame);
end\close(obj);

最后发现是我在循环里面创建 axes 的缘故

改成

ax = axes();%Draw multiple images to display and store each image into a cell array%
for idx = 1:nImages...
end

就好了